Signaling cascades of the Aspergillus fumigatus

2 downloads 0 Views 11MB Size Report
Sep 1, 2013 - DNA damage, extended cell-cycle arrest and differentiation defects52. ..... binding groove (HBG) of another family member or a non-canonical ...
Signaling cascades of the Aspergillus fumigatus virulence factor Gliotoxin in mediating apoptosis and invasive aspergillosis Inaugural-Dissertation zur Erlangung der Doktorwürde der Fakultät für Biologie der Albert-Ludwigs-Universität Freiburg im Breisgau

vorgelegt von Florian Haun geboren in Kassel

angefertigt am Institut für Molekulare Medizin und Zellforschung

Mai 2016, Freiburg im Breisgau

Dekan der Fakultät für Biologie: Prof. Dr. Wolfgang Driever Promotionsvorsitzender: Prof. Dr. Stefan Rotter Betreuer der Arbeit: Prof. Dr. Dr. h.c. Christoph Borner Referent: Prof. Dr. Dr. h.c. Christoph Borner Koreferent: Prof. Dr. Dr. Klaus Aktories Drittprüfer: Dr. Tilman Brummer Datum der mündlichen Prüfung: 02.08.2016

Danksagung

Für die Anfertigung dieser Arbeit habe ich zwischen November 2012 und Mai 2016 dreieinhalb Jahre am Institut für Molekulare Medizin und Zellforschung in Freiburg verbracht. Dies entspricht einer (geschätzten) Arbeitszeit von ca. 7000 Stunden oder 33,600,000 Herzschlägen. Drei Millionen Liter oder ein olympisches Schwimmbecken voller Herzblut also, die in dieses Projekt geflossen sind. An dieser Stelle möchte ich mich bei allen bedanken, die mir dafür die Kraft geben haben. Zunächst möchte ich mich bei meinem Doktorvater Prof. Dr. Christoph Borner bedanken, dafür dass er mich über Umwege in seine Arbeitsgruppe aufgenommen und mir dieses Themas zur Verfügung gestellt hat, mit Freiraum für eigene Ideen. Insbesondere bedanke ich mich für seine immer wieder überraschende Fähigkeit tiefe Depression über unglückliche Daten in puren Optimismus zu verwandeln. Bei Dr. Ulrich Maurer möchte ich mich herzlichst dafür bedanken, dass er mich insbesondere in der Anfangsphase des Projektes, mit viel „frische Küsche“ auf die richtige Spur in den Wirren zellulärer Kinasen gebracht hat. Im Speziellen möchte ich mich bei Katrin Wieland für die herausragende Zusammenarbeit bedanken und dafür, dass Sie immer Herr (oder besser Frau) der diversen to-do Stapel geblieben ist. Des Weiteren gilt mein Dank den Kollegen der Arbeitsgruppen Borner und Maurer, Dr. Simon Neumann, Dr. Lukas Peintner, Laura Faletti, Katharina Mühlbauer, Lisa Schlicher, Prisca Brauns-Schubert und Florian Preiß aka Papa Schubert. Ich bedanke mich bei euch nicht nur für die endlosen wissenschaftlichen Diskussionen und den Input zu meinem Projekt sondern (und insbesondere) für die vielen unvergesslichen Momente abseits der Arbeit. Die gemeinsame Zeit im Schnee, in Kanus, an den Stränden der Welt und in italienischen Pools hat aus Kollegen Freunde fürs Leben gemacht und die Rückschläge im Labor vergessen lassen.

Das im Labor überhaupt gearbeitet werden kann verdanke ich Karin Neubert, Sandra Sandler, Mirela Tataru und Claudia Ortlieb. Ich möchte mich bei euch dafür bedanken, dass ihr nicht nur das Labor am Laufen gehalten habt sondern auch bei jeglichem Problem eine Lösung parat hattet. Des Weiteren gilt mein Dank den Mitgliedern meines Thesis Komitees, Dr. Tilmann Brummer und Prof. Dr. Marco Idzko, für ihre hilfreichen Ideen und Geduld trotz voller Terminpläne. Bei der Spemann Graduate School of Biology and Medicine (SGBM), dem Koordinatoren Team und Kommilitonen, möchte ich mich für die Förderung und Abwechslung zum Alltag bedanken. Ganz besonders möchte ich mich bei meiner Familie und meinen Freunden bedanken, die mich immer wieder aufgebaut und auf Kurs gehalten haben. Es ist schön zu wissen, dass es eine Welt abseits von Westernblot und Caspase assay gibt, die abends auf einen wartet. Vor allen Dingen möchte ich mich bei meiner Freundin Kristine Østevold bedanken. Dafür, dass sie die gesamte Arbeit Korrektur gelesen und bei dem Design der Illustrationen geholfen hat. Vor allem aber dafür, dass sie jeden Tag den Stress der Arbeit mit ihrer Liebe vergessen macht und mich zu jeder Zeit unterstützt und gestärkt hat. Dafür und aus tausend anderen Gründen: Jeg elsker deg masse min kose bams!

Also: Vielen Dank euch allen (und denen die ich vergessen habe) für diese großartige Zeit in der AG Borner und in Freiburg!

Table of contents Abstract ........................................................................................................................ i Zusammenfassung ...................................................................................................... ii

Introduction ................................................................................................................. 1 1

Programmed Cell Death ...................................................................................... 1 1.1

Regulated Necrosis ....................................................................................... 3

1.1.1 1.2

Autophagic Cell Death ................................................................................... 9

1.3

Apoptosis ..................................................................................................... 11

1.3.1 2

2.1

The extrinsic pathway of apoptosis.............................................................. 17

2.2

The intrinsic, mitochondrial pathway of apoptosis ....................................... 20

2.2.1

The Bcl-2 family: regulators of cell death .............................................. 22

2.2.2

Activation of effector proteins and MOMP ............................................. 29

Regulatory networks in cell death ................................................................ 31

2.3.1

Regulation by kinase signaling.............................................................. 31

2.3.2

Clinging to life: cell detachment-induced apoptosis .............................. 37

Fungal infections ............................................................................................... 42 3.1

Invasive pulmonary Aspergillosis................................................................. 43

3.1.1 4

Caspases: executioners of apoptosis ................................................... 13

Orchestration of apoptotic signals ..................................................................... 17

2.3

3

Ferroptosis .............................................................................................. 6

The virulence factor Gliotoxin (GT) ....................................................... 46

Aims of this study .............................................................................................. 50

Materials and Methods ............................................................................................. 52 5

6

Materials ............................................................................................................ 52 5.1

Chemicals and Reagents ............................................................................ 52

5.2

Commercial Kits .......................................................................................... 56

5.3

Inhibitors and Toxins ................................................................................... 57

5.4

Antibodies .................................................................................................... 58

5.5

Plasmids and shRNA vectors ...................................................................... 60

5.6

Primers ........................................................................................................ 61

5.7

Cell Lines ..................................................................................................... 63

Methods............................................................................................................. 65 6.1

Cell Biology ................................................................................................. 65

6.1.1

Cell culture ............................................................................................ 65

6.1.2

Cell death induced by Gliotoxin............................................................. 66

6.1.3

Transfection of pro-apoptotic proteins................................................... 67

6.1.4

Transfection for confocal microscopy.................................................... 67

6.1.5

Lenti- and retroviral infection ................................................................. 68

6.1.6

Cell adhesion assay .............................................................................. 69

6.1.7

Flow cytometry ...................................................................................... 69

6.1.8

Microscopy ............................................................................................ 71

6.2

Biochemistry ................................................................................................ 71

6.2.1

Cell lysis and protein extraction ............................................................ 71

6.2.2

Bradford protein quantification .............................................................. 72

6.2.3

Gel electrophoresis & Immunoblotting .................................................. 73

6.2.4

Immunoprecipitation.............................................................................. 74

6.2.5

Rhotekin pulldown ................................................................................. 75

6.2.6

Mass spectrometry ................................................................................ 76

6.2.7

Caspase activity assay ......................................................................... 77

6.2.8

JNK and ROCK activity assays ............................................................. 77

6.2.9 6.3

Radioactive phosphorylation assay....................................................... 78

Molecular Biology ........................................................................................ 79

6.3.1

Plasmid preparation .............................................................................. 79

6.3.2

Side directed mutagenesis .................................................................... 79

6.3.3

Cloning .................................................................................................. 80

6.4

In vivo experiments...................................................................................... 81

6.4.1

Fungus cultivation ................................................................................. 81

6.4.2

Invasive Aspergillosis model in C57BL/6 mice ...................................... 82

6.4.3

White blood cell counting ...................................................................... 84

6.4.4

Broncho alveolar lavage and lung preparation ...................................... 84

6.4.5

Detection of A.f. infection ...................................................................... 84

6.4.6

Histological sections and fungus staining ............................................. 85

Results...................................................................................................................... 86 7

Bim is released from pro-survival Bcl-2 ............................................................. 86

8

JNK phosphorylates Bim to execute apoptosis ................................................. 87

9

Unraveling the signaling events leading to GT-induced JNK activation ............. 91 9.1

JNK is phosphorylated by MKK4 and MKK7 ............................................... 91

9.2

Inhibition of GSK3 protects from GT-induced apoptosis .............................. 95

9.3

ROCK is a MAPKK kinase for MKK4/7 ........................................................ 98

9.3.1

Cytoskeletal changes during GT-induced cell detachment ................... 98

9.3.2

Pharmacological inhibition of ROCK protects against GT ................... 100

9.3.3

ROCK relays GT signaling to MKK4/7 and JNK.................................. 102

9.3.4

Downregulation of ROCK confirms pharmacological cytoprotection ... 105

9.4

RhoA is activated upstream of ROCK and mediates apoptosis ................. 107

9.4.1 9.5

Inhibition of CDC42 and Rac1 does not protect from apoptosis.......... 111

GT triggers cell death by interfering with cell adhesion signaling .............. 113

9.5.1

GT translocates paxillin from focal adhesions to early endosomes..... 114

10

9.5.2

FAK and p190RhoGAP are both inactivated by GT ............................ 116

9.5.3

GT interferes with the integrin binding capacity to the ECM ............... 122

In vivo validation of GT-induced apoptosis in invasive aspergillosis ............. 131

10.1

Bak-dependent apoptosis is involved in IA ............................................. 133

10.2

Establishment of Bim deficient C57BL/6 mice as an IA model ............... 137

Discussion .............................................................................................................. 140 11

Execution of apoptosis by phosphorylated Bim ............................................ 140

12

Deciphering the JNK activating signaling cascade ....................................... 144

12.1

GSK3 enhances the apoptotic response to GT ...................................... 144

12.2

Gliotoxin employs RhoA to trigger a JNK activating kinase cascade ..... 146

13

The role of integrins and FAK in the anoikis signaling induced by GT .......... 150

13.1 14

GT as a model to study anoikis .............................................................. 157

Apoptosis-deficient mice as a model for invasive aspergillosis (IA) .............. 159

14.1

Inhibition of apoptosis, a possible treatment for IA? ............................... 161

Bibliography ............................................................................................................ 164

Appendix................................................................................................................. 187 15

Supplementary data ...................................................................................... 187

16

Erklärung ...................................................................................................... 190

17

Publications .................................................................................................. 191

18

Abbreviations ................................................................................................ 192

Abstract

Abstract Gliotoxin (GT) is the major virulence factor of the fungus Aspergillus fumigatus (A.f.) and necessary to promote invasive aspergillosis (IA) in immunocompromised patients. Disease progression requires GT-mediated loss of epithelial barrier function to allow the invasiveness of A.fumigatus. GT-induced apoptosis of epithelial cells could contribute to such function. Here we further characterize the apoptotic signaling induced by GT. We show that GT triggers RhoA activation, which induces a kinase cascade via the Rhoassociated kinase (ROCK) and the MAPK kinases MKK4/7. MKK4 and MKK7 activation precedes phosphorylation of c-Jun N-terminal kinases (JNK) and subsequent phosphorylation of Bim. Phosphorylated Bim is released from antiapoptotic Bcl-2 and interacts with Bak to execute GT-induced apoptosis. In this context, ROCK functions as a MAPKK kinase to relay apoptotic signaling in addition to its known function on the cytoskeleton. We found that cells were resistant to GT-induced apoptosis if they were either (i) treated with JNK or ROCK inhibitors or (ii) genetically deleted or knocked-down for ROCK-1 or MKK4/7. GT triggers cell death by a cell detachment-induced mode of apoptosis, termed anoikis. We report that GT can directly bind to cysteines in the ligand binding domain of integrin αVβ3. Furthermore, GT rapidly inactivates integrins to trigger anoikis. Integrin inactivation resulted in the inhibition of focal adhesion kinase (FAK) and p190RhoGAP to activate the RhoA-dependent kinase cascade. We establish GT as a novel, physiological trigger to study anoikis and integrin signaling. Both loss of epithelial cells by apoptosis and/or cell detachment could underlie the reduced barrier integrity. Here we show that ROCK inhibitors not only block GT-induced cell death but also maintain the physiological morphology by preventing detachment of bronchial epithelial cells. Pharmacological inhibition of ROCK might therefore be a promising strategy to prevent the invasive potential of A.fumigatus.

i

Zusammenfassung

Zusammenfassung Gliotoxin (GT) ist der zentrale Virulenz Faktor des Schimmelpilzes Aspergillus fumigatus (A.f.) und verantwortlich für die Manifestation von invasiver Aspergillose (IA) in immunsupprimierten Patienten. Die Invasivität von A.f. beruht hierbei auf dem GTinduzierten Verlust der Barrierefunktion des Lungenepithels. GT-induzierte Apoptose von Epithelzellen könnte in diesem Zusammenhang eine wichtige Rolle spielen. Ziel dieser Arbeit war es die GT-abhängigen, pro-apoptotischen Signal Kaskaden zu charakterisieren. GT aktiviert die GTPase RhoA, welche über die nachgeschaltete Kinase ROCK, die folgende Kinase-Kaskade stimuliert: ROCK phosphoryliert die MAPK Kinasen MKK4 und MKK7, welche wiederum JNK phosphorylieren und somit aktivieren. Diese Kaskade endet mit der JNK-abhängigen Phosphorylierung von Bim und der damit verbundenen Freisetzung von Bcl-2. Bim ist dann in der Lage mit dem Effektor Protein Bak zu interagieren, um die GT-induzierte Perforation von Mitochondrien und somit Zelltod umzusetzen. In diesem Zusammenhang fungiert ROCK als eine pro-apoptotische MAPKK Kinase,

zusätzlich

zu

seiner

bekannten

Funktion

in

der

Regulation

von

Zellmorphologie. Dementsprechend wurde gefunden, dass Zellen resistent gegen GT induzierte Apoptose waren, wenn sie (i) mit JNK oder ROCK Inhibitoren behandelt wurden, oder (ii) ROCK oder MKK4/7 genetisch reduziert beziehungsweise ausgeschaltet waren. Interessanterweise bewirkt GT das Ablösen von Zellen. Dieser Vorgang kann den programmierten Zelltod einleiten und wird dann als Anoikis bezeichnet. Wir konnten zeigen, dass GT direkt relevante cysteine in der Ligandenbindungsdomäne von Integrin αVβ3 modifiziert. Zusätzlich inaktiviert GT Integrine, um Anoikis zu vermitteln. Integrin Inaktivierung induziert die Inhibition der Focal Adhesion Kinase (FAK) und des nachgeschalteten Faktors p190RhoGAP, um die RhoA abhängige Signal Kaskade zu stimulieren. Somit wirkt GT als ein neuartiger und physiologischer Stimulus, um Anoikis zu induzieren und Integrin-abhängige Signaltransduktion zu studieren.

ii

Zusammenfassung Wir zeigen, dass ROCK Inhibitoren sowohl den GT-induzierten Zelltod als auch das Ablösen von Zellen verhindern konnten und daher in der Lage waren, die Epithelzell-Morphologie zu erhalten. Da diese beiden Mechanismen ursächlich für die GT vermittelte Invasivität von A.f. sein könnten, könnte die pharmakologische Inhibition von ROCK eine interessante Strategie zur Bekämpfung von IA darstellen.

iii

Programmed Cell Death

Introduction 1 Programmed Cell Death Cell death is a fundamental basis of life. It is long established that the development from a fertilized embryo to an adult animal requires both generation and death of cells. Programmed cell death has been described by Sulston to be an essential part of tissue development1. In a multicellular organism cells are determined to one of the following fates during development: cell division, differentiation or cell death. The balanced regulation of these processes is substantial for tissue homeostasis and multicellular life in general. An adult human of 80 years would have accumulated an estimated 16 km of intestine and 2 tons of bone marrow and lymph nodes without cell death2. Cell death is classically organized into two different modes. Regulated apoptosis and unregulated necrosis3. More recently a third mode emerged, the autophagic cell death. These three modes of cell death are characterized by distinct morphological features (Figure 1.1). Apoptosis (type I cell death) is characterized by plasma membrane blebbing and nuclear fragmentation. Autophagic cell death (type II cell ceath) can be observed by cytoplasmic vacuolization and loss of organelles. Necrosis (type III cell death) is categorized by rupture of the plasma membrane and organelle swelling. Autophagic cell death and apoptosis are both regulated by genetic programs within the cell. Necrosis however is considered to be an accidental, unregulated response to rigorous stress. The Nomenclature Committee on Cell Death (NCCD) recently suggested to further classify the programmed modes of cell death into five different groups. (1) extrinsic apoptosis, (2) intrinsic apoptosis, (3) regulated necrosis, (4) mitotic catastrophe or mitosis, and (5) autophagic cell death4,5. These classifications are under constant revision and have to be modified continually. Therefore only those modes of cell death with physiological relevance will be discussed here.

1

Programmed Cell Death

Figure 1.1 | Programmed Cell Death. The three major pathways of cell death are shown. Apoptosis (type I cell death), autophagic cell death (type II cell death) and unregulated necrosis (type III cell death). These three types can be distinguished by unique morphological features. Apoptosis is accompanied by nuclear compaction and subsequent fragmentation as well as plasma membrane blebbing. Autophagic cell death is characterized by vacuolization of cytoplasmic components (autophagosome formation) and loss of cellular organelles. Necrosis is denominated by plasma membrane rupture and cytoplasmic leakage. Adopted from6.

2

Programmed Cell Death

1.1 Regulated Necrosis Necrosis can often be seen in response to mechanical forces, heat or cold exposure and hypoxic conditions such as stroke6. Necrosis and subsequent loss of membrane integrity can induce an immune response by the release of damage-associated molecular-pattern (DAMP) molecules7. This early membrane rupture is typically used to discriminate necrosis from apoptosis because cells become permeable for DNA dyes like propidiumiodide (PI) at early stages. In addition, depletion of ATP and increased intracellular calcium concentrations are considered to be hallmarks of necrosis6. Necrosis is classically seen as a random response to injury or rigorous stress. Recent findings showed that receptors that drive inflammatory cytokine production can also induce necrotic cell death. Research on this receptor-linked mode of necrosis changed the accidental, unregulated view of necrosis substantially. Initial work by Holler et al. identified the receptor-interacting protein kinase 1 (RIPK1 or RIP1) as

an

essential

factor

in

tumor

necrosis

factor

(TNF)

family-induced,

caspase-independent cell death8. Furthermore RIP1 kinase activity was needed to execute necrosis, but this activity is dispensable for the classical activation of nuclear factor kappa-light-chain-enhancer of activated B cells (NFκB)8. The dependency of necrosis on kinase signaling was substantiated by the development of a chemical inhibitor, Necrostatin-1 (Nec1). Nec1 is able to block necrotic cell death by inhibiting RIP1 kinase9,10. This form of necrosis is therefore molecularly regulated and has been termed necroptosis by Degterev et al. in 20059. Today regulated necrosis collectively describes several non-apoptotic, regulated modes of cell death that morphologically resemble necrosis. These modes include necroptosis, ferroptosis, pyroptosis, mitochondrial permeability transition and many more11. This chapter focuses on the most intensively studied forms, necroptosis and ferroptosis. RIPK1 is a well-known regulator of NFκB, independent of its kinase activity. It is therefore reasonable that additional proteins might be required in order to specifically activate programmed necrosis. Indeed, RIPK3 was identified by several groups as an essential regulator of regulated necrosis downstream of RPIK112–14. RIPK1 and -3 interact via their RIP homotypic interaction motif (RHIM) forming large amyloid structures, the so called necrosome15,16.

3

Programmed Cell Death Execution of RIPK3 dependent necroptosis was shown to be regulated by activation of mixed lineage kinase like (MLKL) and subsequent production of reactive oxygen species (ROS)13,17–19. How MLKL triggers necroptosis remains elusive11,20. However, it seems that MLKL is phosphorylated by RIPK3, resulting in the formation of MLKL oligomers at the plasma membrane21–23. These oligomers then may serve as a platform for the recruitment of ion channels or they themselves act as a pore forming complex leading to plasma membrane rupture11. Necroptosis can be induced by several stimuli, including members of the tumor necrosis factor family like TNFα, FasL and tumor necrosis factor related apoptosis inducing ligand (TRAIL). A common feature of all necroptosis stimuli is the absence or inhibition of caspases, and therefore necroptosis is clearly different from apoptosis. Among these stimuli, the TNF receptor (TNFR) mediated induction of regulated necrosis is the best studied. TNFR stimulation is usually not associated with cytotoxic cell fate but the formation of a pro-inflammatory/ survival signaling complex (complex I, Figure 1.2)24. This complex has recently been reviewed in detail by Silke and Brink25. In brief, RIPK1 is ubiquitylated by cellular inhibitors of apoptosis (cIAPs) and the linear ubiquitin chain assembly complex (LUBAC), recruiting the TGF activated kinase 1 (TAK1). TAK1 subsequently stimulates NFκB and mitogen activated protein kinase (MAPK) signaling via the IκB kinase (IKK) complex. This stimulation results in the expression of proinflammatory and pro-survival genes25. Inhibition of this NFκB-mediated pro-survival response by destabilization of the receptor bound complex I can lead to a cytosolic, pro-apoptotic complex (complex IIa, Figure 1.2)26. The complex II b or ripoptosome like complex can form in the presence of cIAP inhibitors like second mitochondria-derived activator of caspases (SMAC) mimetics or inhibition of TAK111. The ripoptosome contains RIPK1 and -3 as well as caspase-8 and FLICE-inhibitory protein (FLIPL). Caspase-8 activity is reduced by heterodimerization with FLIPL. The caspase-8-FLIP heterodimer inhibits RIP activity by proteolytic cleavage. However, if caspase activity is blocked or RIPKs are over-expressed, the ripoptosome gives rise to a necroptosis inducing complex, the necrosome11,20.

4

Programmed Cell Death

Figure 1.2 | TNF receptor induced necroptosis. TNF stimulation of the TNF receptor triggers the recruitment of RIPK1 and the formation of complex I. RIPK1 is ubiquitylated by cIAPS and LUBAC in the receptor-bound complex I. Lysine63 ubiquitylation recruits TAK1 and induces subsequent activation of NFκB and MAPK signalling. TNF treatment is therefore considered to be pro-inflammatory and pro-survival by the induction of respective genes. Destabilization or inhibition of complex I leads to the formation of cytosolic, pro-apoptotic complexes (complex IIa and IIb). RIPK activity is proteolytically controlled by caspase-8-FLIP heterodimers in the complex IIb, the ripoptosome like complex. The necroptosis compenent necroptosome or complex IIc forms in the presence of caspase inhibtors. The necroptosome is dependent on RIPK activity and results in the phosphorylation of MLKL. Phosphorylated MLKL forms oligomers at the plasma membrane to induce necrotic cell death. Adopted from11. In summary, most cells are not sensitive to TNF induced cell death. The induction of necroptosis by TNF requires the additional, combinatorial treatment with TAK1 inhibitors or SMAC mimetics together with caspase inhibitors to allow for the formation of the necrosome, if RIPK1 and -3 are expressed at sufficient levels11. This premise makes the physiological relevance of necroptosis highly controversial.

5

Programmed Cell Death There is some genetic evidence that necroptosis might be physiologically relevant. Additional knock-out of RIPK3 in Caspase-8 or Fas-associated protein with dead domain (FADD) deficient mice rescues the embryonic lethality of the latter mice27,28. Moreover, necroptosis is discussed as a key player in the defence of viral infections. RIPK1-RIPK3 dimers are induced in response to vaccinia virus infection and RIPK3-/- mice showed defects in the inflammatory response to the virus13. In addition, viral inhibitors of necroptosis (vIRA) were identified29. Inhibition of necroptosis might be beneficial for viral replication in order to keep cells alive. The fact that viruses like the murine cytomegalovirus evolved inhibitors against necroptosis clearly argues in favor of a physiological contribution of necroptosis, at least in the defence against viruses. More recently necroptosis was linked to a variety of human diseases including ischemia-reperfusion injury, stroke, neurodegeneration, inflammatory diseases and acute kidney injury as reviewed by Jouan-Lanhouet et al.20. Genetic deletion of RIPK3 or inhibition of RIPK1 protected mice from pancreatic oncogenesis30. The ambivalent function of RIPKs in metabolism, apoptosis, inflammation and necrosis, both in a kinase-dependent and -independent way challenges findings based on RIPK knockout studies. The best model to study necroptosis in vivo will be the employment of MLKL deficient mice20.

1.1.1 Ferroptosis A new form of non-apoptotic, non-necroptotic mode of cell death with necrotic features has recently been described31. This form of cell death was shown to be dependent on cellular iron, ROS production and subsequent lipid peroxidation31,32. Therefore it was termed ferroptosis by Dixon et al. in 201231. It is speculated, that ferroptosis is an ancient mode of cell death that developed early during evolution33 but was only recently discovered. Poly-unsaturated lipids are highly vulnerable for oxidation33. Increasing oxygen concentrations in the earth’s atmosphere therefore increased selection pressure towards a defence mechanism against lipid peroxidation. This defence mechanism has developed over time to a complex control mechanism, regulating lipid peroxidation.

6

Programmed Cell Death The key player in this control mechanism is the lipid repair enzyme glutathione peroxidase 4 (GPX4)34. GPX4 is a seleno protein with a glutathione dependent peroxidase activity, thereby protecting against lipid peroxidation. GPX4 activity is crucial for cell survival and is regulated by several pathways (Figure 1.3). Incorporation of a selenocysteine in the active site of GPX4 is regulated via the mevalonate pathway33. This pathway controls the biogenesis of selenocysteine tRNA by the production of isopentenyl pyrophosphate35. In addition, CoenzymeQ10 is produced by the mevalonate pathway. CoenzymeQ10 is an antioxidant, protecting against lipid peroxidation. However, glutathione (GSH) levels represent the main regulation of GPX4 activity33. GSH is produced from cysteine. cysteine can either be synthesized by the transsulfuration pathway or be taken up using antiporters such as the Xccomplex (Figure 1.3)31. The core components of the Xc- complex and GPX4 have been found by screening small molecule libraries31,32,34,36. These studies identified two classes of ferroptosis inducers. Xc- complex inhibitors like Erastin32,36 and GPX4 inhibitors like RSL334. Erastin is toxic by decreasing cystine/cysteine import which results in decreased levels of GSH31. In addition Dixon et al. identified Ferrostatin-1 as a potent ferroptosis inhibitor31. These findings together clearly argue in favor of a regulated mode of cell death. This mode of cell death is not dependent on caspase- or RIPK activity37. It is therefore reasonable to consider ferroptosis as a distinct mode of cell death, although the exact contribution of iron remains elusive33. The physiological relevance of ferroptosis remains controversial. Future studies using GPX4 knock-out mice and the ferroptosis inhibitor Ferrostatin-1 will be needed to clarify the contribution of ferroptosis during development and pathology. First data connects ferroptosis to a variety of human diseases as reviewed by Yang and Stockwell33, including Huntington’s disease. Mutant Huntingtin colocalized with lipid peroxides in striatal neurons of mice38. Mutant Huntingtin protein might therefore be able to kill neurons via the induction of ferroptosis. In addition ferroptosis was recently suggested to be a crucial contributor to p53 dependent tumor suppresion39. This study shows that p53 drives ferroptosis via suppression of Xc- complex components. Mutant p53 that is not able to induce apoptosis or senescence still serves as a tumor suppressor39. It is therefore tempting to speculate that induction of ferroptosis by inhibition of Xc- complex might be beneficial as an anti-cancer treatment33.

7

Programmed Cell Death

Figure 1.3 | Regulation of GPX4 activity and ferroptosis. Ferroptosis is an iron dependent, non-apoptotic but regulated mode of cell death. Iron dependent production of ROS results in lipid peroxidation and subsequent necrotic cell death. Lipid peroxidation is counteracted by the lipid repair enzyme GPX4. GPX4 activity relies on the presence of Glutathione (GSH). GSH levels are reduced during ferroptosis either by reduced synthesis of cysteine via the transsulfuration pathway or by impaired cysteine uptake through the Xc- system. In addition GPX4 activity is regulated by the mevalonate pathway which is required for the incorporation of selenocysteine in the active site of GPX4. Ferroptosis inducers (red) either act on cysteine import (Erastin, Sorafenib), GSH synthesis (buthioninesulfoxamine, BSO) or by directly inhibiting GPX4 activity (RSL3, FIN56). Ferroptosis inhibitors like Ferrostatin-1 (blue) act as antioxidants to protect lipids from peroxidation. Adopted from33

8

Programmed Cell Death

1.2 Autophagic Cell Death Autophagy freely translates into “self-eating” and describes a process in which a cell tries to recycle damaged, toxic or non-essential cytoplasmic components. These components can include everything from misfolded proteins to entire organelles. They are taken up into double-membrane vesicles, the autophagosome6. Autophagosomes are fused to lysosomes in order to degrade the ingested components. Thereby the cell creates a pool of metabolic substrates under nutrient depletion. Consequently autophagy was initially considered to be a pro-survival catabolic process5. Autophagy is a highly regulated multi-step process (Figure 1.4), controlled by approximately 30 autophagy related genes (ATG)5. Some evidence has accumulated, that autophagy may contribute to cell death. However, this autophagic cell death remains controversial40,41. The NCCD suggests to refer to autophagic cell death only from a functional perspective, that is, if cell death can be blocked by inhibition of the autophagic pathway or by genetic depletion of autophagy modulators such as ATG-5, -7 or -12 (Figure 1.4)4,5,42. This classification should help to clearly discriminate autophagic cell death from cell death that is accompanied by autophagy. Nevertheless, it seems reasonable that excessive autophagy beyond a point of cellular rescue can indeed actively induce cell death. The bivalent role of autophagy on cell death has raised research interest over the last couple of years. Inhibition as well as activation of autophagy represent interesting strategies for cancer therapy. Inhibition of autophagy might sensitize cancer cells for apoptosis43,44 whereas induction of autophagic cell death maybe beneficial in apoptosis deficient cancer entities45,46. Many cancer drugs are associated with autophagic cell death as reviewed by Fulda and Kögel5.

9

Programmed Cell Death

Figure 1.4 | Autophagy regulation. Autophagy can be triggered by stress conditions like nutrient depletion, protein aggregation and cancer drugs. Autophagosome formation is regulated by autophagy proteins (ATG). Vesicle initiation is mediated by Beclin-1 (BECN1 or Atg6) dependent activation of Vps34 in the BECN1 core complex. Vps34 is a lipid kinase, creating PtdIns3P which triggers initiation of membrane formation from ER membranes. Elongation and autophagosome formation requires two protein complexes, Atg-5, -12 and -16 and Atg-3, -7 mediated LC3II generation. Mature autophagosomes fuse with lysosomes to form the autolysosome followed by subsequent degradation of the enclosed cargo. Whether autophagy is a pro-survival or cell death inducing mechanism highly depends on the cellular context, the stimulus and the degree of autophagy. The term autophagic cell death should only be used if cell death can be blocked either by inhibition of BECN1 core complex, depletion of the elongation complexes containing Atg5 or Atg7 or by preventing acidification of autolysosomes. Adopted from5,47.

10

Programmed Cell Death

1.3 Apoptosis The term apoptosis was first coined by Foxton Kerr in 197248 and is derived from an ancient Greek word that suggests “leaves falling from a tree”. This metaphor is appropriate in the sense that the programmed sacrifice of single cells is necessary to maintain the functionality of a higher organism. Early molecular research of apoptosis was led by Sydney Brenner, John Sulston and Robert Horvitz. They established the nematode Caenorhabditis elegans (C. elegans) as a model organism for neuronal development and identified the genetic regulation of apoptosis in C. elegans1,49,50. This work paved the way for the discovery of the corresponding genes in humans51. Later it was shown that apoptosis is not only a process to remove damaged cells but also healthy cells during development. It is now accepted that apoptosis is involved in the sculpting and deletion of structures, regulation of cell numbers and the elimination of potentially dangerous cells (Figure 1.5)52. A prominent example for structure sculpting is the formation of limbs. In this process, apoptosis represents the major mechanism to remove interdigital webs53. Furthermore apoptosis is the driving force to delete unwanted or transiently needed structures like the tadpole tail in amphibians, larval tissues in insects, or super numerous cells in sexual organs of humans as reviewed by Jacobson et al.54. In many organs cells are produced in excess and subsequently removed by apoptosis. The nervous system is a prime example for the regulation of cell numbers in that fashion. Barres and Raff estimated that half of all neurons are eliminated by programmed cell death to match their numbers to the amount of cells they innervate55. In addition to the removal of healthy cells in tissue homeostasis, apoptosis also represents a protective mechanism to remove damaged or autoreactive cells. The selection process of B and T lymphocytes by apoptosis is crucial to generate cells that express a functional but not autoreactive antigen receptor52,54,56. Another mode of host protection by apoptosis can be seen in the induction of cell death upon viral infection, DNA damage, extended cell-cycle arrest and differentiation defects52. In summary, it becomes obvious that the ability to induce apoptosis is a key feature for multicellular life. Indeed, defects in the induction of the apoptotic program are a hallmark of cancer development57. On the other hand, over activation of apoptosis is equally harmful and associated with neurodegenerative disorders like

11

Programmed Cell Death Alzheimer’s, Parkinson’s and Huntington’s disease (reviewed by Vila and Przedborski58). An extensive overview of the clinical implications of dysregulated apoptosis is given by Hotchkiss et al.6.

Figure 1.5 | Apoptosis during development. Programmed cell death is involved during various aspects of development and homeostasis. (A) Apoptosis regulates formation of proper structures by deleting interdigital webbings. (B) Larval structures (purple) are eliminated during Drosophila metamorphosis. Functional structures develop from imaginal discs. (C) Apoptosis is the opposing mechanism to cell division regulating cell numbers and tissue homeostasis. (D) Abnormal or dangerous cells are deleted by apoptosis. These cells include autoreactive lymphocytes as well as damaged cells e.g. by DNA damage. Adopted from 52.

12

Programmed Cell Death Morphologically, apoptosis is characterized by cell shrinkage, nuclear condensation (pyknosis) and extensive plasma membrane blebbing59. This is followed by the formation of apoptotic bodies whose membranes and organelles stay intact. On the molecular level apoptosis is associated with the translocation of phosphatidylserine (PS) from the inner leaflet of the plasma membrane to its extracellular site. This exposure is a hallmark of apoptosis60. Thereby, Macrophages recognize and engulf apoptotic bodies and clear them without leakage of any cytoplasmic components. Apoptotic cell death is therefore considered to be non-inflammatory61. The translocation of PS can also be used to mark apoptotic cells for analysis. In the presence of Calcium, fluorescently labelled annexin-V recognizes PS and stains apoptotic cells. However, FITC-annexin-V also marks necrotic cells, because it can be taken up through the perforated plasma membrane and stains PS from the inside. Therefore, quantification of apoptotic and necrotic cells by FACS analysis is done by co-staining with PI, which can only enter necrotic cells (double stain), but not apoptotic cells (single stain). A key event for the execution of apoptosis is the cleavage and activation of caspases. Caspase activity is therefore considered to be a discriminator between apoptotic and necrotic cell death62.

1.3.1 Caspases: executioners of apoptosis Caspases are cysteine proteases that cleave substrates specifically after aspartate residues (Cys asp protease). Once activated, they execute the apoptotic program, orchestrating the morphological changes of apoptotic cell death. Caspases are expressed in healthy cells as inactive precursors (zymogens) that need to be activated to exert their proteolytic function63,64. At least 17 caspases are expressed in mammalian cells, however only a subset of them seem to be involved in cell death40. Non-apoptotic functions of caspases have been observed in the context of inflammatory cytokine processing as reviewed elsewhere64. The apoptotic caspases are classified into two categories, the initiator and the executioner caspases (Figure 1.6, a)65. The initiator caspases (caspase-2, -8 and -9) represent the apical level in caspase activation and are required to activate executioner caspases. The initiator caspases possess a relatively limited spectrum of substrates

13

Programmed Cell Death compared to executioner caspases. They are described to have the potential of selfcleavage, cleavage of executioners and certain minor regulatory proteins of apoptosis such as Bid65. Initiator caspases are usually present as inactive monomers in the cytosol of cells40. The monomers are recruited to various activation platforms where they form dimers. Close proximity of dimeric caspases in these platforms induces conformational changes that can result in their auto-cleavage and the formation of stable, active tetramers that consist of two small and two large subunits (Figure 1.6, b)66,67. The recruitment to activating platforms requires the homophilic interaction of caspases with adaptor proteins in the platform via death effector domains (DED) or caspase activation and recruitment domains (CARD). Initiator caspases harbor these interaction domains at their N-termini (large pro-domain)40. In contrast, executioner caspases are present as inactive dimers. They rely on the activating cleavage by initiator caspases because they do not possess any proteininteraction DED or CARD domains (Figure 1.6, b). Proteolytic cleavage between the large and small subunit creates tetramers with two active sites, representing the mature executioner caspase68,69. Once activated, executioner caspases start to dismantle the cell by cleaving a huge variety of substrates, resulting in the morphological changes associated with apoptosis. These changes are extensively reviewed by Tayler et al. and include the following cleavage events64. Components of the cytoskeleton like myosin, actin and nuclear lamins are cleaved, which contributes to the rounding shape of early apoptotic cells70. Cytoskeletal changes are further increased by cleavage of the Rho-associated kinase (ROCK). Cleaved ROCK is constitutively active and phosphorylates the myosin light chain, resulting in the contraction of actin bundles. These changes further support cell rounding and the formation of membrane blebbs (Figure 1.7). The cleavage of nuclear lamins together with actin contraction tears the nuclear envelope apart, promoting nuclear fragmentation71. In addition, executioner caspases cleave substrates that are linked to transcription like NFκB and eukaryotic initiation factors (eIF’s)72. These events, together with fragmentation of the Golgi and ER, turn off the protein biosynthesis in apoptotic cells (Figure 1.7). Furthermore, the Chromatin is being condensed and degraded. DNA degradation is performed by caspase activated DNase (CAD) in

14

Programmed Cell Death apoptotic cells73. Chromatin condensation, however, is regulated by the caspase activated mammalian sterile-20 kinase (MST1). Another important event in this context is the cleavage of poly (ADP-ribose) polymerase (PARP). PARP is a DNA repair enzyme which is inactivated upon cleavage. Cleaved PARP is routinely used as a readout for caspase activation during apoptosis.

Figure 1.6 | Caspase classification and activation. Cys aspartic acid proteases (caspases) are expressed in cells as inactive precursors which can be activated by proteolytic cleavage. (A) Caspases are categorized into two classes, the initiator and executioner caspases. Both share the presence of a large and a small subunit. Initiator caspases contain an additional pro-domain, the death effector domain (DED). (B) The DED domain allows inactive monomers to localize to caspase activating platforms via adaptor proteins that, in turn, contain a DED. Close proximity in these platforms results in dimerization and auto-cleavage, generating the stabilized, mature protease. Active initiator caspases cleave inactive dimers of executioner caspases, leading to active dimers and execution of apoptotic stimuli. Adapted from65.

15

Programmed Cell Death

Figure 1.7 | Demolition of cellular organelles by caspases. Active caspases execute cell death by inhibition of several cellular life-lines. Caspases directly cleave cytoskeletal components like myosin and actin. Proteolytic activation of ROCK induces actin contraction, resulting in the rounding of cells and the formation of membrane blebbs. Protein biosynthesis is inhibited on several levels. DNA is condensed and fragmented by MST1 and CAD. Transcription factors (NFκB) and translation initiators (eIF’s) are cleaved and inactivated. Additionally active caspases fragment the ER and Golgi. Finally, focal adhesions and cell adhesion sites are targeted to facilitate phagocytosis of apoptotic cells. Adapted from64. Altogether, caspases target almost 400 different substrates regulating different aspects of cellular survival (Figure 1.7)72. It is difficult to discriminate between apoptosis triggering cleavage events and by-stander substrates due to the unspecific nature of caspase activity. However, it seems obvious that caspase-induced cell death results from cleavage of many cellular proteins, rather than one specific target64.

16

Orchestration of apoptotic signals

2 Orchestration of apoptotic signals Most apoptotic stimuli converge on the activation of caspases. A cell is doomed to die once caspases are activated. Hence, the decision to activate caspases is tightly regulated. Apoptotic pathways can be classified depending on the mode of initiator caspase activation. Activation via death receptors (DR) at the plasma membrane is a hallmark for the extrinsic pathway of apoptosis. Death receptors are activated by extracellular ligands of the TNF superfamily which are classically expressed by neighboring cells. In contrast, many cellular stresses like DNA damage or growth factor withdrawal induce a different pathway of apoptosis that involves the permeabilization of mitochondria. This pathway is known as the intrinsic pathway because the cell fate decision is made within the cell.

2.1 The extrinsic pathway of apoptosis Cell death execution by the extrinsic pathway emanates from death receptors (DR) at the membrane. All DR share the presence of a cytosolic death domain (DD). DR are activated by formation of homo-trimers after ligand binding. The ligands are expressed by cells of the immune system and belong to the TNF superfamily, including TNFα, Fas ligand (FasL or CD95L) and TRAIL. Trimerization of the respective receptors induces accessibility of their DD in the cytosol74. The receptor DD is subsequently recognized and bound by DDs of adaptor proteins, especially FADD. Receptor bound FADD represents the core of a complex known as the death inducing signaling complex (DISC)75. Interactions at the DISC are based on homotypic interactions between the different death domains. FADD contains a DED in addition to its death domain, allowing further interactions with DED containing proteins. These proteins include pro-caspase-8 and -10 as well as the FLIP proteins (Figure 2.1, a). Both pro-caspases as well as FADD contain additional DED domains, allowing for either homo- or hetero dimerization as exemplified by the activation of CD95 (Figure 2.1, a). A recruited pro-caspase can either recruit another pro-caspase, leading to close proximity induced proteolytic activation or recruit an inhibitory FLIP protein, restraining apoptotic capacity76,77. The regulation of these interactions dictates apoptotic outcome in response to CD95L stimulation. In that sense CD95 activation

17

Orchestration of apoptotic signals was reported to be either pro-apoptotic or able to induce non-apoptotic pathways like NFκB or MAPK signaling, depending on a threshold of CD95L78,79. The activation of caspases was long believed to be a simple two-step process with one dimer per FADD (Figure 2.1, b)74. Caspase-8 homo-dimerization first results in a cleavage of the small protease subunit, generating a p43/p41 subunit bound to FADD via its DED and a p10 subunit that binds to the p18 subunit of the caspase. Thereby, both products remain associated to the DISC. In a second cleavage event the large protease domain is cleaved from the DED containing pro-domain. This cleavage produces the active p18/p10 tetramer74,76.

Figure 2.1 | Apoptotic signaling at the DISC. Pro-caspases are activated by death receptors (DR) at the plasma membrane. (A) Ligands of the TNF family induce trimerization of DR subunits. Ligand binding exposes the cytosolic death domain (DD), recruiting the adaptor protein FADD via homotypic interaction with the FADD death domain. FADD binding induces the accessibility of its DED for binding of Pro-caspase8 or -10 or FLIP. FLIP and pro-caspases form dimers via DED-DED interaction generating either non-apoptotic heterodimers or cell death inducing homo-dimers of caspase-8. (B) Caspase-8 is activated in two proteolysis steps induced by close proximity. The first cleavage occurs between the large and the small protease subunit, creating a p43/41 subunit and a p10 fragment which is bound to the p18 subunit. The second cleavage releases the active tetramer to induce executioner caspases. Adopted from74.

18

Orchestration of apoptotic signals Two independent studies now suggest that the stoichiometry of DISC signaling is more complex than initially assumed. These studies identified up to 9 caspase molecules bound to the DISC for every FADD80,81. This “caspase chain model” can only be explained if the DED of FADD can recruit more than one DED protein82. Indeed caspase-8 and FLIP possess differential affinity for binding surfaces exposed by the DED of FADD83. FLIP binds to a binding site defined by α1/α4 helices of FADD DED and caspase-8 preferentially binds to a surface generated by α2/α5 (Figure 2.2) 83. This study further modified the caspase chain model. They could not confirm the sub-stoichiometric presence of FADD (personal communication, DB Longley, 2015), but identified the formation of FLIP-FADD-caspase trimers via the differential binding of FLIP and caspase-8 to FADD. These trimers are able to oligomerize, forming signaling competent chains (Figure 2.2)83. Pro-caspase-8 remains unprocessed at the DISC if FLIP(S) binds to the α1/α4 surface of FADD. The incorporation of FLIP(L) instead of (S) results in the formation of a membrane-restricted, non-apoptotic signaling complex with low protease activity. High levels of caspase-8 or FLIP depletion can result in the formation of trimers with caspases bound to both interfaces of FADD. Incorporation of such a trimer in the chain results in the formation of pro-caspase-8 homodimers and proximity induced proteolytic activation as described above. The formation of active caspase-8 tetramers executes the extrinsic apoptosis signaling by activating the executioner caspases-3/-6 or -7. This direct activation (type I cell-death pathway) of executioner caspases is sufficient to induce cell death in cells with high DISC formation and therefore high levels of active caspase-8. However, some cells like hepatocytes (type II cells) have low DISC formation capacity. In these cells caspase-8 triggers an amplification mechanism by activating the intrinsic pathway via cleavage of the protein Bid.

19

Orchestration of apoptotic signals

Figure 2.2 | Caspase chain model. Binding of FADD to the DR exposes two interfaces on the FADD-DED, α1/α4 and the α2/α5 interface (rear and front of the chevron, respectively). FLIP proteins (yellow) bind preferentially to the α1/α4 interface and procaspase-8 (violet) to the α2/α5 interface, forming trimers of FLIP-FADD-Caspase-8. Incorporation of FLIPshort creates an inactive signaling complex. FLIPlong containing trimers generate membrane bound complexes with low proteolytic activity. Low levels of FLIP can yield trimers containing pro-caspase-8 on both FADD interfaces. Interaction of these trimers results in proximity induced activation of pro-caspase-8, forming the apoptosis competent tetrameric caspase-8. Caspase-8 cleaves executioner caspases to transmit the apoptotic stimulus. Adopted from83.

2.2 The intrinsic, mitochondrial pathway of apoptosis The intrinsic pathway is the most common mode of programmed cell death6. It is also termed the mitochondrial pathway because the activation of caspases is controlled by mitochondrial integrity. Cellular stress regulates the Bcl-2 protein family to induce the permeabilization of the mitochondrial outer membrane (MOMP). MOMP releases mitochondrial factors that help to activate the most important initiator caspase of the intrinsic pathway, pro-caspase-9. These pro-apoptotic proteins include cytochrome c and Smac/DIABOLO84,85. Smac is able to antagonize the inhibitor of apoptosis (IAP) protein family. IAPs can inhibit either the initiator caspase9 via its BIR3 domain or the executioner caspases via its BIR2 domain86,87. Smac release is therefore an amplifying apoptotic event. The crucial step for activation of pro-caspase-9 is the release of cytochrome c and the subsequent formation of the apoptosome. The apoptosome acts as a cytosolic

20

Orchestration of apoptotic signals activation platform of caspase-9, consisting of 7 molecules of the apoptotic protease activating factor 1 (Apaf-1)88. In healthy cells Apaf-1 resides in an auto-inhibitory conformation with the c-terminal WD40 repeats intramolecularly associated. Thereby, Apaf-1 is locked as a cytosolic monomer (Figure 2.3, b)87,89. Cytochrome c release unlocks the auto-inhibited monomer by binding to the WD40 repeats. Apaf-1 then forms a heptameric wheel-like structure in an ATP-dependent process. The caspase recruitment domains (CARD) of Apaf-1 form a ring structure in the center of the apoptosome upon oligomerization90.

Figure 2.3 | Apoptosome formation. The apoptosome is a heptameric, wheel-like structure consisting of Apaf-1, cytochrome c and caspase-9. (A) The apoptotic protease activating factor 1 consists of an N-terminal CARD domain followed by an ATPase domain of the AAA+ family and c-terminal WD40 repeats. The AAA+ family of ATPases is known to be activated by oligomerization. (B) Apaf-1 is present as an autoinhibitory monomer in healthy cells. MOMP and subsequent cytochrome c release opens the locked conformation of Apaf-1. Apaf-1 forms a heptameric wheel upon ADPATP exchange to recruit and activate the initiator caspase-9 via its centralized CARD domains. Adopted from87. Exposure of the Apaf-1 CARD domains recruits inactive, monomeric caspase-9 to the apoptosome via its own N-terminal CARD domains. The initiator caspase-9 is most likely activated by dimerization in the apoptosome66,91. The active, dimeric

21

Orchestration of apoptotic signals initiator caspase cleaves the inactive precursors of the executioner caspases-3 and -7 to implement the apoptotic stimulus89. As discussed above, the oligomerization of Apaf-1 and subsequent activation of caspase-9 requires the release of cytochrome c and an additional ATP-dependent conformational change of Apaf-1. Furthermore, IAPs have to be inhibited in order to promote cell death. It is reasonable to speculate that this regulation on different levels has evolved in order to prevent accidental caspase activation due to cytochrome c leakage87.

2.2.1 The Bcl-2 family: regulators of cell death MOMP is considered to be the “point of no return” of intrinsic cell death92. Once permeabilized, cytochrome c is released from the mitochondria and triggers the activation of caspases and subsequent dismantling of the cell. The decision of MOMP execution is therefore tightly regulated by the Bcl-2 family. The oncogene B-cell lymphoma-2 (Bcl-2) is the founder of the Bcl-2 family93. In contrast to other oncogenes studied during that time, it did not promote cell proliferation but was found to inhibit cell death94. Today it is well established, that the Bcl-2 family orchestrates apoptotic signaling and MOMP. The members of this family can be divided into pro-apoptotic and anti-apoptotic members. All members share the presence of one to four Bcl-2 homology domains (BH). The pro-survival, anti-apoptotic Bcl-2 proteins are characterized by four BH domains and include Bcl-2 itself, Bcl-xL, Bcl-w, myeloid cell leukemia 1 (Mcl-1), A1 and Bcl-B95. The pro-apoptotic members are further divided into the multi-domain effector proteins Bcl-2 associated X protein (Bax), Bcl-2 antagonist killer 1 (Bak), Bcl-2 related ovarian killer (Bok) which consist of four BH domains and the BH3-only proteins, which harbor only the BH3 domain. The BH3-only proteins include Bcl-2 interacting mediator of cell death (Bim), Bcl-2 interacting domain death agonist (Bid), p53-upregulated modulator of apoptosis (Puma) and others as depicted in Figure 2.4, a. The BH3-only proteins can be additionally divided into direct activators of Bax and Bak and derepressor/sensitizer proteins of Bax/Bak activation (Figure 2.4, a). Yet, the direct activators can also function as derepressors95–98.

22

Orchestration of apoptotic signals The effector proteins Bax, Bak, and possibly also Bok perform MOMP upon oligomerization. This oligomerization is regulated by protein-protein interactions between the pro- and anti-apoptotic proteins. Interactions within the Bcl-2 family are governed by canonical binding of the BH3 domain of one protein to the hydrophobic binding groove (HBG) of another family member or a non-canonical binding outside the HBG.

Figure 2.4 | The Bcl-2 family interactions. (A) Schematic representation of the Bcl2 family members, classified as direct activator BH3-only proteins, derepressor BH3onlies, multi-domain effector proteins and the anti-apoptotic proteins. The BH domains one to four constitute the structurally well-defined Bcl-2 core, shared by anti-apoptotic proteins and the effector proteins. In contrast, the BH3-only proteins are less structured. (B) Overview of the binding specificities within the Bcl-2 family. Protein interactions are based on the binding of a BH3 domain into the hydrophobic binding groove (HBG) of another family member. The groove is delineated by the helices displayed in red (A). Adopted from95. The hydrophobic, cytosolic exposed binding groove is formed between the helices α2-α5 of the BH3 and BH1 domain, respectively (Figure 2.5, a)99,100. An engaged BH3 domain adopts a helical structure by the formation of up to 7 hydrophobic interactions within the groove, even if the domain originates from an initially unstructured protein (Figure 2.5,b)95,99. The structural features in the groove dictate the differential binding affinities within the Bcl-2 family as reviewed by Moldoveanu et al. (Figure 2.4, b)95.

23

Orchestration of apoptotic signals

Figure 2.5 | Hydrophobic binding groove and BH3 binding. The Bcl-2 core of BclxL. (A) The HBG binds the BH3 domain and the C-terminal transmembrane domain of Bcl-2 family proteins. It is formed by the helices α2 and α5 of the BH3 and BH1 domain, capped by α8. (B) BH3 - HBG interaction, exemplified by Bcl-xL and the BH3 domain of Bad. Conserved amino acids (Leu and Asp) of the BH3 domain engage via hydrophobic and electrostatic interactions to amino acids of the groove. The structure of the HBG defines binding affinities within the family. Adopted from95. Non-canonical interactions target allosteric sites outside the groove. These interactions regulate Bax/Bak oligomerization and modulate the binding affinity of BH3only proteins to the HBG of their respective binding partner. The best characterized non-canonical interactions are described for the activation of Bax by Bim101,102. The combination of canonical and non-canonical interactions defines the complex regulatory network orchestrating the apoptotic response (Figure 2.6). The interactions within the Bcl-2 family are of transient nature and require the presence of membranes, especially the mitochondrial outer membrane. Biochemical data is therefore often questionable and/or difficult to reproduce because binding assays are performed in solution or with various detergents. This resulted in the proposal of many different interaction models such as the “neutralization model”103,104, the “direct

24

Orchestration of apoptotic signals activation model”105,106 and the “embedded together model”107. Llambi et al. tried to assemble these scenarios to a unified model of intrinsic apoptosis (Figure 2.6)96. The anti-apoptotic Bcl-2 proteins either inhibit the direct activator proteins tBid, Bim and Puma or antagonize the effector proteins Bax/Bak at the mitochondrial membrane (embedded together)95,96. Derepressor BH3-only proteins such as Bad, Bmf, and Hrk exert their pro-apoptotic activity by binding to and inhibiting anti-apoptotic proteins, thereby relieving the effector proteins Bax and Bak from inhibitory complexes by competing for binding sites (neutralization model). In addition a subset of the BH3-only proteins (tBid, Bim, Puma) has been shown to directly stimulate Bax/Bak oligomerization (direct activator model)106. The unified model includes all of these possibilities as depicted in Figure 2.6. The exact mechanism of execution and contribution of the different family members can be different, depending on the cell type and nature of the stimulus.

Figure 2.6 | Orchestration of apoptotic stimuli. Pro-apoptotic oligomerization of the effector proteins can be stimulated by direct-activator BH3-only proteins and is antagonized by anti-apoptotic proteins. In addition anti-apoptotic proteins can inhibit direct activators and in turn, direct activators can release Bax/Bak from inhibitory complexes with anti-apoptotic Bcl-2 proteins. Similarly, derepressor BH3-only members can induce the accessibility of effector proteins for direct activators by binding to anti-apoptotic proteins. Adopted from40. 25

Orchestration of apoptotic signals The dependency of regulatory interactions on the HBG-BH3 interface has recently been exploited in the context of cancer therapy. The ability to escape from apoptosis is a hallmark of cancer cells and contributes to chemo resistance. Anticancer drugs often induce cell death in tumor cells by activation of BH3-only proteins via upstream signaling modules, like p53. However, such BH3-inducing mechanisms are often mutated, silenced or lost in tumor cells, rendering them resistant to apoptosis108,109. A new class of compounds has thus been developed in the effort to adjust the Bcl-2 driven regulation of cell death in order to restore apoptosis sensitivity. These drugs resemble the BH3 domain and bind to the HBG of selected anti-apoptotic proteins. These BH3-mimetics like ABT-737 or its analog ABT-263 selectively bind to and inhibit the pro-survival proteins Bcl-2, Bcl-xL and Bcl-w (but not Mcl-1 or A1). Inhibition of these proteins releases BH3-only proteins from inhibitory complexes to inactivate Mcl-1 (sensitizer proteins) or to directly activate Bax/Bak110– 112.

Under physiological conditions, the Bcl-2 family members are regulated by transcriptional and post-transcriptional mechanisms, formation of multi-protein complexes and by a variety of posttranslational modifications. Transcriptional activation is exemplified by the induction of Puma and Noxa by p53 in response to DNA damage113,114. Different miRNA were reported as post-transcriptional regulators of both pro- and anti-apoptotic Bcl-2 family members, as reviewed by Lima et al.115. Phosphorylation of BH3-only proteins represents another major strategy of apoptosis regulation. Anti-apoptotic phosphorylation of Bad via AKT sequesters and inactivates Bad in the cytosol by binding to 14-3-3 proteins116. Pro- and anti-apoptotic phosphorylation also regulates Bim activity, which will be discussed in the following chapter.

26

Orchestration of apoptotic signals 2.2.1.1 The BH3-only protein Bim Bim was first described in 1998 by O’Connor et al. and was reported to be expressed in 3 isoforms termed BimEL (extra-long) BimL (long) and BimS (short), according to their molecular weights117. All isoforms share the BH3 domain, located in exon 5, whereas only BimEL and BimL display a dynein light chain interaction site118. The regulation and contribution of these isoforms to cell death remains elusive. However, Bim is established as one of the most potent direct activators of Bax and Bak with an additional high binding affinity to all anti-apoptotic proteins40,95,97,98. Physiological importance of Bim first emerged in the context of lymphoid and myeloid cell homeostasis. Bim knock-out mice develop autoimmune diseases due to defects in the clearance of auto-reactive thymocytes119. Furthermore, Bim was shown to be required for apoptosis induced by UV stress, cytokine withdrawal and glucocorticoids120. These findings raised interest to discover how Bim is regulated. Subsequent studies revealed that the Bim activity is controlled by transcriptional and posttranslational mechanisms. Bim levels are increased in response to cytokine withdrawal by the transcription factor FOXO3a, whereas ER-stress induces Bim via CHOP121,122. Several miRNA clusters are described to post-transcriptionally regulate Bim in mice as well as in human cancer cell lines108. In addition, Bim activity is known to be regulated by posttranslational phosphorylation. In healthy cells extracellular signal-regulated kinase-1/2 (ERK1/2) phosphorylates S55/S65/S73 of BimEL, priming it for proteasomal degradation and thereby keeping Bim at low levels 123,124. In addition, Bim phosphorylation has been proposed to regulate its interaction with pro-survival proteins and multi protein complexes including the dynein light chain (DLC)125–127. It was long believed that DLC interaction recruits unphosphorylated Bim to cytoskeleton complexes in healthy cells, keeping Bim inactive. However, more recently it was shown that Bim is constitutively located at mitochondria128 and that DLC rather facilitates the formation of inactive Bim multimers. BimEL and BimL possess a dynein light chain interaction site which allows the formation of Bim dimers, oligomers and high molecular weight complexes at mitochondria, which most likely keep Bim in check in healthy cells (unpublished data, Daniel Frank, Häcker Lab).

27

Orchestration of apoptotic signals

Figure 2.7 | Regulation of pro-apoptotic Bim. High levels of Bim antagonize prosurvival proteins like Mcl-1 or directly activate Bax/Bak to perforate the mitochondrial outer membrane. Bim levels are regulated by survival signaling pathways either transcriptionally via inhibition of FOXO3a or by direct phosphorylation. ERK dependent phosphorylation of Bim at S55/S65/S73 induces proteasomal degradation of Bim, promoting cell survival. In contrast triple phosphorylation of Bim at Ser100, Thr112 and Ser114 induces the pro-apoptotic potential of Bim. Modified from129. In contrast to the protective ERK phosphorylation, Bim can also be phosphorylated by c-Jun N-terminal kinases 1/2 (JNK1/2) upon different apoptotic stimuli, including UV exposure130. Bim phosphorylation at position Thr112 by JNK was reported to disrupt the interaction of Bim with DLC131. It is therefore tempting to speculate, that JNK mediated phosphorylation releases Bim from high molecular weight complexes at mitochondria. In line with this, the JNK phosphorylation site overlaps with the DLC binding site of Bim130. In that sense, JNK phosphorylation would release monomeric Bim, inducing the antagonization of anti-apoptotic Bcl-2 family members or direct activation of the effector proteins Bax/Bak. Surprisingly, the Bim T112A mutation only partially protects cells from apoptosis131,132. Consequently we could show that the apoptotic activity of Bim is induced by a triple phosphorylation of Bim by JNK at Ser100, Thr112 and Ser114132. 28

Orchestration of apoptotic signals

2.2.2 Activation of effector proteins and MOMP The effector proteins Bax and Bak regulate mitochondrial membrane integrity downstream of the anti-apoptotic proteins and the direct activators. The exact activation of Bax and Bak remains elusive and is considered the “holy grail” of apoptosis research97. Bax and Bak might have redundant functions in the regulation of MOMP. Mice deficient of both proteins, but not the single knock-out mice, exhibit impaired apoptosis during development53. Cells derived from these mice fail to induce MOMP, underlining the crucial role of Bax and Bak in this process133. Despite some similarities, essential differences are present regarding the localization of these proteins. Bak is mainly localized to mitochondria via its C-terminal trans-membrane domain. Bax also contains a membrane targeting domain (α9) but it is buried in its hydrophobic groove, and Bax is therefore predominantly found in the cytosol of healthy cells134. Bcl-xL was recently shown to shuttle Bax back into the cytosol once it engages mitochondria, thus preventing oligomerization and MOMP, in a process termed Bax retro-translocation135. The same group could later show that also Bak retro-translocates to the cytosol. The shuttling rates of Bak are so low that it appears to be mitochondrial but can be found, to some extent, in the cytosol of human tissues and cells136. Bax and Bak contain a hydrophobic BC groove (HBG) similar to the antiapoptotic proteins. This supports the direct activator model in which Bax requires the binding of BH3 peptides in the groove to be activated. However, recent data suggest that Bax (and Bak) might also be activated by non-canonical interactions by binding of BH3-onlies to a hydrophobic surface on the opposite site of the protein134,137,138. However mutational studies of this interface have questioned these findings139. Activation of Bax and Bak induces extensive conformational changes that result in mitochondrial targeting and homo-oligomerization65,106. These changes have recently been reviewed extensively134. Structural analysis of truncated Bax suggested the core-latch domain model of Bax activation (Figure 2.8)137,138,140. Monomeric, inactive Bax translocates, possibly facilitated by non-canoncical binding of BH3-only proteins, to the mitochondrial outer membrane which induces the exposure and subsequent integration of the α9 helix in the membrane, resembling the native Bak. The α9 displacement renders the hydrophobic groove accessible for BH3only activator proteins. BH3-groove interaction induces the formation of a destabilizing 29

Orchestration of apoptotic signals cavity in Bax which promotes the disengagement of the core (α2-α5) and the latch domain (α6-α8)137. This disengagement leads to the release of the initiator BH3 domain and exposure of the Bax/Bak BH3-domain. In this state the BH3 domain of one effector protein can bind to the exposed groove of a neighboring effector to form symmetrical homo-dimers. These dimerized Bax/Bak molecules serve as the building block for subsequent oligomerization (Figure 2.8). The symmetric dimers most likely interact via a planar surface composed of α4 and α5134,137

Figure 2.8 | Conformational changes during Bax and Bak activation. The corelatch model by Czabotar et al.138. (1) Bax retro-translocates to the mitochondrial membrane, eventually aided by non-canonical binding of BH3 proteins. The α9 transmembrane helix is released from the hydrophobic groove and inserted in the membrane. (2) BH3-only proteins can engage the exposed hydrophobic groove of Bax/Bak. (3) BH3-only:groove interaction induces an internal cavity that disengages the latch from the core domain. (4) Latch disengagement results in the release of the activating BH3-only protein and exposure of the Bax-BH3 domain. (5) Bax/Bak forms symmetric homo-dimers by binding of one BH3 domain into the groove of another effector protein. (6) Dimers can further oligomerize in order to perforate the outer mitochondrial membrane, possibly via a hydrophobic surface created by α4 and α5. Adopted from134.

30

Orchestration of apoptotic signals The exact structure and composition of the Bax/Bak pore remains elusive. Two possible architectures have been proposed. The formation of proteinaceous pores, channels that consist only of oligomerized Bax or Bak, and lipidic pores that are stabilized by Bax and Bak are being discussed65. The employment of super-resolution microscopy recently revealed membrane pores, delineated by Bax oligomers141. In this study the authors suggest that 8-10 Bax oligomers might be sufficient to stabilize 100 nm wide pores that also contain lipids. They conclude with a lipidic pore model where the amount of incorporated Bax regulates the diameter of the pore141. The third family member of the effector proteins, Bcl-2 related ovarian killer (Bok) is by far the less well characterized. It has only recently been shown that endogenous Bok indeed exerts pore formation capacity in the absence of Bax and Bak142. In contrast to Bax and Bak, Bok is not regulated by the proteins of the Bcl-2 family but is constitutively active. Instead, Bok is rapidly degraded via the endoplasmicreticulum-associated degradation pathway (ERAD) and therefore kept at low levels in healthy cells142.

2.3 Regulatory networks in cell death 2.3.1 Regulation by kinase signaling Kinases are key regulators of cellular functions and are involved in the transmission of pro-apoptotic signals to the Bcl-2 family. Kinase signaling controls the activity of Bcl-2 proteins by a variety of regulatory mechanisms. These mechanisms act on both antiand pro-apoptotic Bcl-2 family members and include the transcriptional control of protein levels via transcription factors, targeting of proteins for proteasomal degradation and the modulation of protein-protein interactions between family members40. The phosphoinositide-3-kinase (PI3K) - Akt pathway represents a prominent pathway to control of cell survival and cell death. Akt (also known as protein kinase B) is recruited to the membrane by PI3K-dependent formation of phosphatidylinositol (3,4,5)-triphosphate (PIP3) in response to receptor tyrosine kinases, G-protein coupled receptors or integrins. Akt binds PIP3 via its pleckstrin homology domain (PH) and is then activated by other kinases like PDK1143. Active Akt regulates cell death via the

31

Orchestration of apoptotic signals forkhead boxO 3a (FOXO3a) transcription factor, modulation of other kinases and phosphorylation of BH3-only proteins like Bad. FOXO3a is sequestered and inhibited by 14-3-3 proteins in the cytosol in response to phosphorylation by Akt. Therefore the expression of pro-apoptotic proteins like Bim and Puma, as well as death receptor ligands like TRAIL and FasL, are inhibited40,143. Similarly, Akt phosphorylates Bad (Ser112/136) which is inactivated by 14-3-3 proteins in the cytosol116. In addition, Akt regulates protein levels of Mcl-1 via the glycogen synthase kinase 3 (GSK3). Growth factor deprivation inactivates PI3KAkt signaling, relieving GSK3 from the inhibitory phosphorylation by Akt. Once unleashed, GSK3 induces cell death via upregulation of Puma and degradation of Mcl-1. GSK3 phosphorylates and activates the histone acetyl transferase TIP60. TIP60 stabilizes p53 to induce the expression of Puma144. GSK3 also induces the proteasomal degradation of Mcl-1 by phosphorylation and subsequent ubiquitylation, to enhance the apoptotic signaling145. Interestingly, GSK3-dependent phosphorylation of Mcl-1 requires a priming phosphorylation by c-Jun n-terminal kinases (JNK)146. JNK belongs to the superfamily of mitogen activated protein kinases (MAPK). The three major members of this family are JNK, p38 and extracellular-signal-regulated kinases (ERK). MAPKs relay various signals from the plasma membrane and are considered to be key regulators of cellular functions, including gene expression, proliferation, mitosis, motility, metabolism and apoptosis. Activation of MAPKs follows a strict hierarchy (Figure 2.9)147. Signals from growth factor receptors or G proteincoupled receptors are transduced via MAPK kinases (MAPKK) which in turn are activated by another tier of kinases, the MAPKK kinases (MAP3K). This process is usually facilitated by scaffold proteins. Activation of ERK is associated with pro-survival outcome and is reported to activate c-myc. In addition, ERK keeps the pro-apoptotic protein Bim at low levels by inducing Bim proteasomal degradation via phosphorylation or by inhibition of Bim expression via phosphorylation of FOXO3a (see 2.2.1.1). In contrast, JNK and p38 are activated in response to environmental stresses and share overlapping MAPKKs (Figure 2.9). Therefore p38 is often stimulated simultaneously with JNK. The outcome of p38 activation is highly dependent on the

32

Orchestration of apoptotic signals context and has been described to be both pro-survival and cell death inducing (reviewed by Wada et al.147). Especially CD4+ T-cells seem to require p38 activation by its MAPKK MKK-6 (but not MKK-3) to execute cytokine withdrawal-induced apoptosis148. The underlying mechanisms however remain elusive147.

Figure 2.9 | MAPK signaling. MAPKs are serine/threonine kinases that are activated by dual phosphorylation. The hierarchical activation of MAPKs is mediated by specific MAPKKs like MKK4 and MKK7 for JNK. MAPKKs are activated by a more diverse spectrum of MAPKKKs with partial overlap. ASK1 and TAK1 for example, can activate MAPKKs for p38 and JNK. Together the MAPKs relay signals from the plasma membrane, regulating virtually all aspects of cellular life. Adopted from abcam Biochemicals. (http://docs.abcam.com/pdf/biochemicals/MAPK-pathway.pdf, 06.04.16). The role of JNK in the induction of cell death can be divided into nuclear and mitochondrial signaling149. Initially, JNK was described to phosphorylate c-Jun in the nucleus in response to UV stress

150.

Phosphorylated c-Jun forms hetero dimers with

c-Fos to form the transcription factor AP1. AP1 dependent transcription has been reported to induce the expression of pro-apoptotic proteins like FasL, TNFα and Bak149,151. The importance of nuclear JNK-AP1 signaling was confirmed by the

33

Orchestration of apoptotic signals employment of non-phosphorylatable c-Jun mutants. These mutants render MEFs resistant to UV irradiation152. Nevertheless, JNK also interacts with other transcription factors like the activating transcription factor 2 (ATF2), ETS domain-containing protein (Elk-1), p53 and c-myc147. The overall outcome of nuclear JNK is therefore highly dependent on the nature, the strength and the duration of the stimulus153. In addition, JNK is known to directly modulate the proteins of the Bcl-2 family. As described above for Bim (chapter 2.2.1.1), JNK phosphorylates Bim at three sites to increase its pro-apoptotic capacity. Similarly, phosphorylation of Bad by JNK at Ser128 promotes apoptosis in rat neurons (Figure 2.10)154. In contrast to Akt mediated phosphorylation, Bad phosphorylated at Ser128 by JNK is released from inhibitory interaction with 14-3-3 proteins to antagonize anti-apoptotic Bcl-2 proteins155. Furthermore, JNK was described to inactivate pro-survival proteins. The JNK-priming phosphorylation of Mcl-1 for GSK3 phosphorylation and subsequent degradation promotes cell death signaling, as described above. In addition, Bcl-2 can be inactivated by JNK-mediated phosphorylation at Ser70 during cell cycle progression156 and Bcl-xL is neutralized by phosphorylation at Thr47/115157. Although these studies established a crucial role of JNK in the regulation of mitochondrial apoptosis, the exact mechanisms of how these phosphorylations execute apoptosis remain elusive. The conformational changes and respective changes in binding affinities remain to be determined. The Rho-ROCK signaling axis is another crucial kinase pathway in the modulation of cell death and survival. The Rho family belongs to the Ras superfamily of small GTPases and includes the Rho proteins, Rac and CDC42. Small GTPases are activated by guanosindiphosphate (GDP) to guanosintriphosphate (GTP) exchange and are subsequently inactivated by hydrolysis of GTP. This switch function is dictated by activating guanine nucleotide exchange factors (GEFs) and Rhoinactivating GTPase activating proteins (GAPs). Active RhoA stimulates ROCK by displacing the C-terminal Rho binding domain from an auto-inhibitory conformation. ROCK is an essential kinase in the regulation of the actin myosin cytoskeleton, cell adhesion, cell motility and polarity158. Phosphorylation of LIM kinase or myosin light chain phosphatase (MLC) by ROCK inhibits actin polymerization and promotes cell contractility159. In addition, ROCK activity is also related to the induction and execution of cell death (Figure 2.11).

34

Orchestration of apoptotic signals

Figure 2.10 | JNK signaling in apoptosis. c-Jun n-terminal kinases are activated by the MAPKKs MKK-4 and -7 in response to several stimuli. Active JNK contributes to apoptosis via nuclear translocation and transactivation of transcription factors. These transcription factors induce the expression of apoptotic proteins like FasL or Bak. JNK also acts directly on the mitochondrial pathway. Phosphorylation of Bad at Ser128 releases it from 14-3-3 complexes to antagonize Bcl-2. Phosphorylation of Bim increases its pro-apoptotic potential to directly activate Bax and Bak or to inhibit prosurvival proteins. The apoptotic response is amplified by JNK-dependent inhibition of pro-survival proteins like Bcl-2. Modified from149. In contrast to ROCK’s influence in cell survival and contractility, the apoptotic signaling is less well understood but might affect the extrinsic as well as the intrinsic pathway159. ROCK activity is essential for the clustering of the FAS receptor via the phosphorylation of adaptor proteins that link the receptor to the cytoskeleton160,161. The intrinsic pathway is affected by inhibition of the PI3K/Akt pathway which might translate into transcriptional regulation of Bcl-2 family members159. The ROCK activity on the cytoskeleton explains reported effects on the formation of membrane blebs, nuclear membrane- and Golgi fragmentation162,163.

35

Orchestration of apoptotic signals

Figure 2.11 | ROCK functions in apoptosis. The Rho-associated kinase ROCK is activated by binding of RhoA to its C-terminal rho-binding domain. ROCK is a key regulator of the cytoskeleton and can induce cell contractility. Therefore, ROCK facilitates the phenotypic changes of apoptotic cells, like membrane blebbing and nuclear membrane fragmentation. In addition, ROCK can signal to the intrinsic pathway by inhibition of PI3K/Akt and companions the clustering of the FAS receptor during activation. Adopted from159. The exact role of ROCK in apoptosis remains to be characterized, especially regarding the question if ROCK is a regulator or rather an executioner of apoptosis.

36

Orchestration of apoptotic signals

2.3.2 Clinging to life: cell detachment-induced apoptosis The attachment of cells to the extracellular matrix (ECM) via integrins is a vital process. Integrins not only mediate the physical contact of the cytoskeleton to the surrounding environment, but also relay signals from the ECM to the cell.164. The integrin family consists of eighteen α and eight β subtypes that form at least 24 different transmembrane heterodimers at the plasma membrane. Alpha subtypes are usually composed of a seven blade β-propeller representing the ligand binding head domain and a leg domain with one thigh and two calf domains (Figure 2.12). The β head domain is created by a hybrid domain with an inserted β-I domain. The β-I domain further harbors a metal ion-dependent adhesion site (MIDAS) for ligand binding. The leg of the β subunit is conformationaly more flexible than the α leg and contains four cysteine-rich epidermal growth factor (EGF) modules165. These EGF modules are crucial for the activation of integrins in the transition from a folded conformation to an open confirmation. This process involves the re-shuffling of disulfide bonds that are tied between the EGF domain cysteines166–170. Integrins can function in a bidirectional way. Signals that are created in the cell (inside-out signaling) increase the affinity for the ECM components. On the other hand, binding to the ECM matrix induces intracellular conformational changes which is known as outside-in signaling165. These signaling events have been shown to play a role in cell survival for at least four integrin heterodimers (αVβ3, α1β1, α6β1, α5β1)171. Integrin-mediated survival signaling is regulated via activation of the MAPKs ERK and JNK as well as activation of the PI3K-Akt signaling axis. These proteins trigger genetic programs to promote proliferation and differentiation (Figure 2.13). In addition, these pathways restrain the execution of apoptosis by inhibiting BH3-only proteins,172 as discussed above (chapter 2.3.1).

37

Orchestration of apoptotic signals

Figure 2.12 | Integrin domain structure. Integrins are transmembrane heterodimers comprised of α and β subtypes forming at least 24 different combinations. (D) Each subtype has a head domain for ligand binding, a leg domain for positioning of the head, a transmembrane domain, and a cytosolic domain for signal transduction. Inactive integrins occupy a closed conformation. (A) Activation requires the re-arrangement of disulfide bonds between cysteines of the β subunit’s EGF modules (E1-E4, green) and involves an open transition conformation. (B) Ligand binding induces a further conformational change of the hybrid domain to fully activate integrins. (C) This conformational change is transduced to the cytosolic domain to activate integrin signaling. Adopted from165.

38

Orchestration of apoptotic signals

Figure 2.13 | Integrin signaling during cell survival. Integrins anchor the cytoskeleton to the ECM and relay pro-survival signaling via MAPKs, the PI3K/Akt pathway and small GTPases. FAK is a key protein sensing and transmitting signals from integrins to these pathways, regulating cell morphology, cell cycle, and proliferation

in

response

to

external

stimuli.

Source:

https://mbi-

figure.storage.googleapis.com/figure/1385625597741.jpg, 07.04.2016. Focal adhesion kinase (FAK) is the major switch at focal adhesions to relay signals from the ECM to the survival pathways. ECM engagement and subsequent formation of focal adhesions recruits FAK via the adaptor proteins paxillin and talin to the cytosolic tail of integrins and to the cytoskeleton. Upon this recruitment, autoinhibition of FAK is relieved, leading to auto-phosphorylation at Tyr397, which serves as a binding site for the kinase Src (Figure 2.14). Src fully activates FAK by additional phosphorylations in the activation loop (Tyr576/577)173,174. Together, Src and FAK activate their downstream targets either directly as for PI3K/Akt or indirectly as seen for ERK acitvation175. FAK phosphorylation (Tyr925) by Src generates a binding site for the growth factor receptor-bound protein 2 (Grb2). Grb2 is an adaptor protein to activate the Ras-Raf-MEK-ERK pathway176.

39

Orchestration of apoptotic signals

Figure 2.14 | FAK domain structure and interactions. Focal adhesion kinase contains an N-terminal FERM domain, which allows interactions with the EGF- and platelet derived growth factor receptor (PDGF). The FERM domain binds to the kinase domain in the autoinhibitory complex. Recruitment to focal adhesion adaptor molecules like talin or paxillin is mediated by the C-terminal focal adhesion targeting domain (FAT) and uncages FAK. Subsequent autophosphorylation at Tyr397 recruits Src, which fully activates FAK by phosphorylation in the kinase domain. Additional phosphorylation of Tyr925 in the FAT domain serves as a docking site for p190RhoGAP inactivating RhoA, and for Grb2 activating Ras. Modified from177. In addition, FAK regulates morphological changes via the Rho family of GTPases. An important role for FAK-RhoA signaling was described for the regulation of endothelial barrier function. In this context, active FAK has a crucial role in maintaining barrier integrity by inactivating RhoA via p190RhoGAP178.The ERK activating phosphorylation of FAK at Tyr925 also serves as a docking site for p190RhoGAP, which is activated by FAK mediated phosphorylation178–181. Active p190RhoGAP shuts down RhoA, preventing the permeability of the endothelial membrane. Detachment from the ECM promotes the activation of apoptotic programs. This cell detachment-induced mode of apoptosis is termed anoikis from the Greek word for “homelessness”171. Loss of this anchorage-dependent regulation of cell death is considered a milestone of cancer progression. Anoikis usually makes sure that detached transformed cells undergo apoptosis, thus preventing the survival of tumor cells in suspension and subsequent metastasis. The importance of anoikis inhibition for cancer progression is highlighted by the diverse strategies which cancer cells use to evade detachment-induced cell death. These strategies include the “integrin switch”,

40

Orchestration of apoptotic signals referring to a modified integrin profile on the cell surface, constitutive activation of survival pathways and epithelial-mesenchymal transition171. How anoikis is executed molecularly remains elusive but seems to involve both extrinsic and intrinsic apoptosis signaling pathways171. Most of these effects have been linked to the inactivation of pro-survival PI3K/Akt and ERK signaling by loss of cell anchorage. Inactivation of ERK was shown to repress the expression of anti-apoptotic FLIP along with an induction of FasL182,183. The termination of these pathways in response to detachment was also linked to the intrinsic pathway. Ras-dependent depletion of Bcl-xL was shown to play a causal role in anoikis of intestinal epithelial cells184. Furthermore, MAPK inactivation due to detachment results in increased levels of Bim by transcriptional and posttranslational mechanisms (chapter 2.3.1). Indeed, Bim and also Bid have been implied to play a pivotal role in the induction of anoikis172,185. This notion is substantiated by recent findings that BimEL is sequestered, and therefore inactivated, during anoikis evasion of breast cancer cells186. In summary, it becomes evident that anoikis, at least in part, is executed via the Bcl-2 family and requires the BH3-only protein Bim. However, no clear data is available showing how loss of anchorage is transduced into apoptotic signaling. Anoikis research is hampered by a lack of appropriate models to study detachment-induced cell death which are desperately needed in order to understand how cancer cells evade anoikis. New models will help to identify which molecules could be targeted to resensitize cancer cells for this cell death process. Fighting anoikis resistance is an appealing strategy to target metastatic cancer cells. Therefore, detailed knowledge about the cellular response to detachment and the hierarchy of signaling events will be needed to identify possible targets171.

41

Fungal infections

3 Fungal infections The kingdom of fungi is morphologically classified into unicellular yeast and filamentous forms (molds). Fungi occupy a wide range of habitats and have a huge impact on human life. Applications of natural products from fungi range from antibiotics to the preparation and preservation of food. Despite their beneficial applications, fungi also represent a major threat for human health, species extinctions, food security, and the balance of ecosystems187. Over 600 different fungi species have been reported to infect billions of people every year188. The infections range from minor skin infections to deadly infections, killing as many people as tuberculosis or malaria. Skin and nail infections are by far the most common fungal diseases that affect approximately 25% of the worldwide population189. Invasive infections have a much lower incidence, yet they represent the major threat of fungal infections due to the high mortality rates, killing one and a half million patients every year. The three most significant invasive infections are caused by Aspergillus fumigatus (>200,000 infections/year), Candida albicans (>400,000 infections/year) and Cryptococcus neoformans (>1,000,000 infections/year)189. The ubiquitous exposure to fungal spores and cells requires a fine-tuned immune response of the host. This response can be classified into two components: resistance and tolerance190. The tolerance component describes adaptations of the immune system to prevent host damage due to excessive and permanent immune reaction, reviewed by Romani191. Resistance in this context describes the ability to actively reduce fungal burden. This reaction is mainly regulated by pattern recognition receptors (PRRs) like Toll-like receptors (TLRs) and C-type lectin receptors (CLRs)192. These receptors recognize pathogen-associated molecular patterns (PAMPs) in fungi, especially glycan- and galactomannan structures. Recognition of PAMPs triggers the NFκB dependent secretion of cytokines and induction of phagocytosis by macrophages, leading to direct fungal killing191. The balanced execution of resistance and tolerance enables humans to fight, or at least restrain, fungal infections. Therefore, fungal infections were considered to be of low importance in the early 20th century. The drastic progress and growing importance of immunosuppressive therapies (organ and stem cell transplants, autoimmune

diseases,

chemotherapy)

and

the

increasing

escalation

of

immunosuppressive infections such as HIV/AIDS, however, boosted the susceptibility 42

Fungal infections to fungal infections. Invasive fungal infections account for 50% of all AIDS related deaths193. These circumstances express the pressing need for research on fungal infections and the development of therapeutic possibilities. According to Brown et al., three major aspects have to be addressed in order to tackle these infections188. Robust, specific and cheap diagnosis is needed to implement antifungal treatments. Safer and more effective drugs need to be developed in order to avoid problems with bioavailability, drug resistance, toxicity and high costs of therapy. And finally, vaccines have to be developed to fight allergies and life threating infections before they develop.

3.1 Invasive pulmonary Aspergillosis The ubiquitous mold Aspergillus fumigatus (A.f.) and other related fungi cause a severe pulmonary disease termed invasive aspergillosis (IA or invasive pulmonary aspergillosis IPA) in immunocompromised people. A.f. is a saprophytic mold inhabiting human environments, especially soil and organic debris. The asexual reproduction is based on the conidiophore, a structure that releases airborne conidia. A.f. is characterized by a high sporulating capacity194. Therefore, hundreds of these airborne conidia are inhaled daily by humans. Conidia reach the terminal airways and pulmonary alveoli, owing to their small diameter (2.0 - 3.5 µm)195. Conidia are counteracted in healthy people by alveolar macrophages which destroy conidia upon phagocytosis. In addition, the unique composition of the A.f. cell wall can trigger a receptor mediated immune response. As described above, these receptors are mainly represented by TLRs and CLRs like Dectin 1 and Dectin 2. The secretion of cytokines recruits neutrophils that help to eradicate the fungus from the lung196. Furthermore, it has been reported that the lung epithelial cells can engulf conidia197. However, whether epithelial uptake of conidia represents a pathogenic or host defense mechanism remains eluisve198. It has been speculated that internalized conidia might represent a reservoir for further infection. However, 97% of all conidia are killed within 24 h post phagocytosis, arguing against a pathogenic mechanism199. Additionally, epithelial cells express Dectin 1 and can therefore promote an immune response by releasing cytokines198.

43

Fungal infections A.f. has evolved several strategies to evade the immune system. Conidia and hyphae release superoxide dismutase (SOD), A.f. diffusible product (AfD) and conidial inhibitory factor (conidial IF) to counteract attacks by reactive oxygen species and phagocytosis. Furthermore opsonization by complement and surfactant proteins like surfactant protein A (SP-A), SP-D, soluble IgA and c-reactive protein (CRP) is inhibited by the release of complement inhibitory factors (complement IF)200,201. Finally, secreted proteases might dampen the mucociliary activity of epithelial cells to remove conidia from the lung (Figure 3.2)200. These strategies can explain the fact that prolonged inhalation of conidia can cause allergic responses such as Allergic Bronchopulmonary Aspergillosis (ABPA) in the presence of an intact immune system202. However, the missing immune response in immunosuppressed patients transforms the inhalation of conidia to a life-threating condition. In these patients, conidia can grow unabated and colonize the lung, leading to massive damage of the lung tissue. Unopposed growth finally leads to the accumulation of fungal bulbs, the so called aspergilloma (Figure 3.1)189,194,195.

Figure 3.1 | A.f. growth in the lung. A.f. grows unabated in immunosuppressed patients. Inhaled conidia undergo germ tube formation and subsequent generation of hyphal networks. These networks give rise to a massive, solid fungal structure, the aspergilloma. CT scan of an infected patient showing massive lung damage. Adopted from189.

44

Fungal infections The early symptoms of an A.f. infection resemble those of a regular flu and include fever and cough. The diagnosis of A.f .infections is therefore difficult and often too late. At early stages the fungus can, to some extent, be treated with antimykotica. These anti-fungal drugs either target the cell wall, DNA synthesis or sterol synthesis at the endoplasmic reticulum189. However, the performance of all approved drugs is rather disappointing. The close evolutionary relationship to humans (compared to bacteria) limits the scope of specific drug targets. The available drugs failed to improve the mortality rates of 40-90% due to an unfavorable toxicity profile, undesirable drug interactions with corticosteroids as well as poor bioavailability and absorption. In addition, resistance mechanisms have been described which are partly due to environmental fungicide use189,203. If A.f. is not cleared from the lung, it is able to break the lung epithelial barrier and invade the lung tissue (Figure 3.2). The fungus then enters the blood flow and is distributed to secondary infection sites. The chances for a successful treatment further decrease once A.f. is invasive189. Blocking the invasiveness of A.f. therefore represents an appealing strategy to treat IA. The mechanisms of how the fungus is able to break the lung epithelial barrier remain mysterious. A.f. harbors a variety of toxins, including Gliotoxin (GT), Verruculogen, Fumagillin and helvolic acid, that certainly facilitate the invasiveness during disease progression. Among these, GT was shown to be the most important for the development of IA204. It is therefore important to study the mechanisms of how GT promotes the invasive potential of A.f. in order to improve the treatment of patients.

45

Fungal infections

Figure 3.2 | A.f. breaks the lung epithelial barrier. Inhaled conidia reach the terminal airways. Several factors like SOD, AfD and complement IF are secreted in an attempt to evade the immune response of healthy people. However, in immunocompromised patients A.f. can break the lung epithelial barrier and is distributed to secondary sites by entering the blood flow. The exact mechanisms remain elusive but certainly require the secreted toxin Gliotoxin. Adopted from200.

3.1.1 The virulence factor Gliotoxin (GT) GT is a secondary metabolite of the Aspergillus species and the prototype of the epidithiodioxopiperazine-type of fungal toxins. This class of toxins is characterized by a reactive disulfide bridge that mediates redox activity by repetitive cleavage and restoration of this bridge (Figure 3.3)205,206. The first scientific interest in GT was raised during the establishment of penicillin as a potent antibiotic fungal product. Indeed, GT was shown to have antistatic but not antibiotic properties207,208. Many studies have focused on the activity of GT since its structure was clarified in 1958209. These studies revealed that GT toxicity highly depends on an intact, oxidized disulfide bridge in the molecule201,210,211. It was speculated that the major function in that context is the covalent modification of free cysteines on target proteins, altering protein function (Figure 3.3)210.

46

Fungal infections

Figure 3.3 | Gliotoxin structure and mode of action. Toxicity of the epidithiodioxopiperazine GT is characterized by an intact disulfide bond in its oxidized form (1). It was postulated that this bond allows the covalent modification of free cysteine thiols to exert toxicity. Adopted from210. Gliotoxin biosynthesis is catalyzed in several steps by proteins that are encoded by genes in the Gli-gene cluster (Figure 3.4). Deletion of any of these genes results in decreased or ablated GT production206,212. The high mortality rates of A.f. infections indicate that GT production did not evolve to invade human hosts. Indeed, intensive studies on the biosynthesis pathway revealed that GT might act as an antioxidant to protect the fungus from reactive oxygen species213,214. The protein that mediates the final step in GT synthesis, the GT oxidase GliT, not only oxidizes GT but also confers resistance to exogenous GT. It is tempting to speculate that this protein is able to maintain the intact disulfide bond by constant oxidation, thereby inactivating exogenous GT and conferring protection of A.f. from its own toxin206,215. Furthermore, the production of GT is considered to mediate a benefit over other fungi that compete for nutrients. GT was shown to inhibit growth of pathogenic fungi such as Candida Albicans206,216.

47

Fungal infections

Figure 3.4 | Gliotoxin biosynthesis. The mycotoxin Gliotoxin is synthesized by proteins encoded by the Gli-gene cluster. Initially, two amino acids (Phe and Ser) are condensed by GliP to form the basic structure. GT is formed via reactive intermediates by methylation, degradation and oxidation. Finally, GliT oxidizes the free thiols to form the reactive disulfide bridge. Fully processed GT is exported by GliA. Reduction of the disulfide bridge e.g. by glutathione inactivates GT. Additionally, reduction of GT explains its function as an antioxidant. Adopted from 217 and 206. The toxic effects of GT on other microorganisms and its impact on animal cells remain elusive. Today it is evident that GT is a major contributor to the development of IA and A.f. invasiveness, but how it promotes the invasive potential of A.f. is not as well characterized212. Mice infected with wild-type A.f. (B5233) developed IA between 10 to 30 days, whereas mice infected with a mutant A.f. strain that lacks the GliP gene for GT biosynthesis were protected212. Initially, GT was described to exert immunosuppressive effects by inhibiting NFκB signaling218. Further studies showed that the immunosuppressive capacity also includes the inhibition of antigen-presenting cells219 which in turn prevents T cell activation and cytokine production220. In addition, GT was shown to induce apoptosis of monocytes221. As described above, IA mainly effects immune compromised patients. It is therefore difficult to image that the immunosuppressive properties of GT contribute

48

Fungal infections to the invasiveness of A. fumigatus. Finally, the GT-mediated inhibition of NFκB inhibition was challenged by studies, showing that GT actually stimulates NFκB dependent interleukin-8 production in human bronchial epithelial cells222. Interleukin-8 production and subsequent recruitment of neutrophils might controversially represent a mechanism to increase the invasiveness of the fungus. A.f. is protected from reactive oxygen species (ROS) released from neutrophils by the antioxidant function of GT as well as GT-induced inhibition of the NADPH oxidase assembly that is required in neutrophils to produce ROS223. An interleukin driven inflammation therefore might damage the lung tissue, allowing A.f. to invade the host. This notion is further underlined by the fact that non-neutropenic mouse models are more likely to develop IA than neutropenic models224. A hallmark of GT toxicity is the induction of apoptosis in several cell types including monocytes, fibroblasts, bronchial epithelial cells and colorectal cancer cells221,225,226. Pardo et al. have shown that GT-induced apoptosis is mediated by the Bcl-2 family effector protein Bak. Bak, but not Bax deficient mice were protected from IA in response to A.f. infection226. This study showed that (i) apoptosis is crucial to promote invasiveness and (ii) GT-induced cell death is executed by the mitochondrial pathway of apoptosis. Subsequent work in our laboratory further underlined the importance of the mitochondrial pathway by showing that Bak activation is regulated by JNK-dependent pro-apoptotic triple phosphorylation of Bim132. Mouse embryonic fibroblasts deficient for Bim or JNK were significantly protected, if challenged with GT. In contrast, deletion of other BH3-only proteins did not confer GT resistance. Overall it is clear that GT is able to target several cellular pathways to elicit its toxic effects. The covalent modification of thiol groups by GT could explain the rather unspecific targeting properties of GT and its divers signaling outcomes. Further studies are therefore required to establish the role of GT in IA.

49

Aims of this study

4 Aims of this study Airborne conidia of A.f. are a ubiquitous threat to human health. Humans daily inhale hundreds of these spores that reach the pulmonary alveoli. A.f. is counteracted by alveolar macrophages and neutrophils in healthy people. However, the increasing use of immunosuppressive agents results in a myriad of patients that fail to respond properly to A.f. conidia. Fungal spores can grow unabated in these patients, forming large hyphal networks and aspergilloma, resulting in damage of the lung tissue. The high mortality rates of up to 90%, however, are due to the potential of A.f. to invade the patient and establish a systemic infection. Secondary infection sites include the heart, brain and liver. Once A.f. is invasive, the prognosis for successful therapy is poor. Conventional strategies try to target the fungus directly. However, up to date no effective antifungal drugs have been found to significantly improve the devastating mortality rates. Here, we consider to fight the disease by inhibiting the breach of the lung epithelium. Therefore, the exact mechanisms employed by A.f. to break the lung epithelial barrier have to be elucidated. Gliotoxin is the major virulence factor in this context and is known to affect a variety of cellular pathways. However, preventing the induction of apoptosis might represent the most promising strategy to block A.f. invasiveness since Bak-deficient mice were protected from A.f. infections. Furthermore, we could show earlier, that GT executes apoptosis via JNK. JNK performs a triple phosphorylation of Bim that possibly alters its affinities to pro-survival Bcl-2 proteins and/or the effector protein Bak. Interestingly, GT induces cell detachment before triggering apoptosis, which might contribute to the loss of epithelial barrier function. Therefore, the aims of this study are to: (i)

Elucidate the molecular machinery that activates JNK.

(ii)

Identify the cellular target(s) of GT.

(iii)

Clarify the contribution of cell detachment to GT-induced cell death.

(iv)

Pinpoint possible drug targets to block the invasiveness of A. fumigatus

(v)

Establish in vivo models to verify potential targets.

50

Aims of this study

Figure 4.1 | Aims of this study. JNK executes pro-apoptotic GT signaling by phosphorylation of Bim. GT most likely activates JNK via targets at the plasma membrane. GT can enter the cell by diffusion but is inactivated by glutathione. Cell detachment of cells prior to apoptosis might point to the involvement of integrins. JNK is a kinase sensing cellular stress. Therefore it might be possible that GT induces perturbations of the membrane or binds to other membrane proteins. Adopted from206. An intact disulfide bond is crucial for GT toxicity. Intracellular GT is rapidly reduced and thereby inactivated by glutathione. It is therefore tempting to speculate that GT has to act at proteins located at the plasma membrane, or the membrane itself. JNK is activated by phosphorylation in response to GT. Thus, the strategy is to identify upstream-kinases of JNK that are activated in response to the toxin. These kinases might lead to the apoptosis inducing receptor of GT. 51

Materials

Materials and Methods 5 Materials 5.1 Chemicals and Reagents

Table 5.1 | Used chemicals and reagents Compound

Supplier

25 x complete protease inhibitor

Roche Applied Science, Mannheim

3 x Flag peptide

Sigma-Aldrich GmbH, Taufkirchen

Ac-DEVD-AMC fluorogenic substrate

Axxora LLC, Farmingdale

Acrylamide (30%)

AppliChem GmbH, Darmstadt

Adenosine triphosphate (ATP)

Sigma-Aldrich GmbH, Taufkirchen

Agarose

Lonza GmbH, Wuppertal

Ammonium bicarbonate

Sigma-Aldrich GmbH, Taufkirchen

Ammonium peroxide sulfate (APS)

Carl Roth GmbH & Co, Karlsruhe

Ampicillin

Carl Roth GmbH & Co, Karlsruhe

Annexin-V APC

Borner laboratory, Freiburg

Annexin-V FITC

Borner laboratory, Freiburg

Anti-HA affinity matrix

Roche Diagnostics, Mannheim

Attractene

Qiagen, Hilden

Bis-TRIS

Carl Roth GmbH & Co, Karlsruhe

Bovine serum albumin (BSA)

Sigma-Aldrich GmbH, Taufkirchen

Bradford Protein Solution

BioRad, München

Bromophenol blue

Merck-VWR Int. GmbH, Darmstadt

52

Materials Butyrate

Sigma-Aldrich GmbH, Taufkirchen

Calcium chloride

Merck-VWR Int. GmbH, Darmstadt

CHAPS

Calbiochem (Merck), Darmstadt

Coomassie Blue G250

Carl Roth GmbH & Co, Karlsruhe

Coulter Isotone II Diluent

Beckman Coulter, Krefeld

Crystal violet

Sigma-Aldrich GmbH, Taufkirchen

Cyclo-RGD-5-FAM

AnaSpec, Fremont

Developing solution for Curix 60 (G153) AGFA Health Care, Köln Digitonin

Carl Roth GmbH & Co, Karlsruhe

Dimethyl sulfoxide (DMSO)

Sigma-Aldrich GmbH, Taufkirchen

Disodium hydrogen phosphate

Carl Roth GmbH & Co, Karlsruhe

Dithiothreitol (DTT)

AppliChem GmbH, Darmstadt

DMEM Medium (4.5 g/l glucose)

PAA Laboratories GmbH, Cölb

DNA-ladder (100 bp)

Fermentas GmbH, St. Leon Rot

dNTPs

Fermentas GmbH, St. Leon Rot

EDTA

Sigma-Aldrich GmbH, Taufkirchen

EGTA

Carl Roth GmbH & Co, Karlsruhe

Enhanced chemiluminescence (ECL)

Perbio Science GmbH, Bonn

ERK (recombinant)

ProQinase, Freiburg

Ethanol

Sigma-Aldrich GmbH, Taufkirchen

EZview Red ANTI-FLAG M2 Affinity Gel Sigma-Aldrich GmbH, Taufkirchen Fetal calf serum (FCS)

Biochrom AG, Berlin

Fibronectin (human)

Advanced Biomatrix, San Diego

Fixation solution for Curix60 (G354)

AGFA Health Care, Köln

Formaldehyde

Sigma-Aldrich GmbH, Taufkirchen

53

Materials Gamma P32/33 ATP

Hartmann Analytic, Braunschweig

Gliotoxin

Sigma-Aldrich GmbH, Taufkirchen

Glutamine

Invitrogen, Karslruhe

Glycerol

Carl Roth GmbH & Co, Karlsruhe

Glycin

Carl Roth GmbH & Co, Karlsruhe

Go Taq G2

Promega, Mannheim

HEPES

Sigma-Aldrich GmbH, Taufkirchen

Human rec. Integrin αVβ3

R&D systems, Minneapolis

Hydrocortisone

Sigma-Aldrich GmbH, Taufkirchen

Interleukin-3

Peprotech, Rocky Hill

Imidazol

Sigma-Aldrich GmbH, Taufkirchen

Iodacetamide

Sigma-Aldrich GmbH, Taufkirchen

Isopropyl alcohol

Fisher Scientific, Loughborough

JNK2 (recombinant)

ProQinase, Freiburg

Kanamycin

Sigma-Aldrich GmbH, Taufkirchen

Lambda Protein Phosphatase

New England Biolabs

Leupeptin

Alexis Deutschland GmbH, Grünberg

Lipofectamine 2000

Lifetechnologies, Carlsbad

Lyticase

Sigma-Aldrich GmbH, Taufkirchen

Magnesium chloride

Merck-VWR Int. GmbH, Darmstadt

Magnesium sulfate

Carl Roth GmbH & Co, Karlsruhe

Mannitol

Sigma-Aldrich GmbH, Taufkirchen

Methanol

Merck-VWR Int. GmbH, Darmstadt

Milk powder, blocking grade

Carl Roth GmbH & Co, Karlsruhe

54

Materials Neomycin (G-418)

Lifetechnologies, Carlsbad

Nonident P-40

Fermentas GmbH, St. Leon Rot

Opti-MEM

Invitrogen GmbH, Karlsruhe

PageRuler Prestained Protein Ladder

Fermentas GmbH, St. Leon Rot

Paraformaldehyde (PFA)

Carl Roth GmbH & Co, Karlsruhe

Penicillin/ Streptomycin (P/S)

PAA Laboratories GmbH, Cölb

Pepstatin

Sigma-Aldrich GmbH, Taufkirchen

Peptone

Sigma-Aldrich GmbH, Taufkirchen

Phosphatase Inhibitor Cocktail Set IV

Merck-VWR Int. GmbH, Darmstadt

PMSF

Sigma-Aldrich GmbH, Taufkirchen

Polybrene

Sigma-Aldrich GmbH, Taufkirchen

Ponceau S

Sigma-Aldrich GmbH, Taufkirchen

Potassium chloride

Carl Roth GmbH & Co, Karlsruhe

Potassium Dihydrogen Phosphate

Merck-VWR Int. GmbH, Darmstadt

Potassium Hydroxide

Merck-VWR Int. GmbH, Darmstadt

Propidium Iodide (PI)

Sigma-Aldrich GmbH, Taufkirchen

Puromycin

Sigma-Aldrich GmbH, Taufkirchen

RHNaseH

Invitrogen GmbH, Karlsruhe

Rhotekin-RBD beads

Cytoskeleton Inc., Denver

RPMI-1640

PAA Laboratories GmbH, Cölb

Saponin

Sigma-Aldrich GmbH, Taufkirchen

S.O.C medium

Invitrogen GmbH, Karlsruhe

Sabouraud Agar

Merck-VWR Int. GmbH, Darmstadt

Sodium azide

Sigma-Aldrich GmbH, Taufkirchen

55

Materials Sodium carbonate

Sigma-Aldrich GmbH, Taufkirchen

Sodium chloride

Merck-VWR Int. GmbH, Darmstadt

Sodium dihydrogen phosphate

Merck-VWR Int. GmbH, Darmstadt

Sodium dodecyl sulfate

Carl Roth GmbH & Co, Karlsruhe

Sucrose

Sigma-Aldrich GmbH, Taufkirchen

SYBR green Safe

Invitrogen GmbH, Karlsruhe

TEMED

Sigma-Aldrich GmbH, Taufkirchen

Thiopental

Inresa GmbH, Freiburg

Tris-HCl

Sigma-Aldrich GmbH, Taufkirchen

Triton X-100

Sigma-Aldrich GmbH, Taufkirchen

Trizol

Invitrogen GmbH, Karlsruhe

Trypsin-EDTA

Lifetechnologies, Carlsbad

Tween-20

Merck-VWR Int. GmbH, Darmstadt

5.2 Commercial Kits

Table 5.2 | Molecular Biology tool kits Kit

Supplier

Accustain Silver Stain

Sigma-Aldrich GmbH, Taufkirchen

DuoLink II proximity ligation assay

OLink Bioscience, Uppsala

Gel extraction Kit

Qiagen, Hilden

Human IL-8 ELISA DuoSet

R&D systems, Minneapolis

Jetstar Plasmid Purification Kit

Genomed GmbH, Löhne

KinaseSTAR JNK activity assay Kit

BioVision Inc., Milpita

56

Materials Platelia Aspergillus ELISA

BioRad, München

SilverQuest SilverStaining Kit

Thermo Fisher, Waltham

Wizard Plus Miniprep DNA Purification Kit

Promega, Mannheim

Quick Change Site directed mutagenesis Kit

Agilent Technologies, Waldbronn

ZymoPure Maxiprep Kit

ZymoResearch, Freiburg

5.3 Inhibitors and Toxins

Table 5.3 | Inhibitors and Toxins Inhibitor

Target

Final conc.

Supplier

Axon-1126

GSK3

0.75 µM

Axon MedChem, Hanzeplein

C2II/C3FT

Rho

200/ 100 ng/ml Aktories laboratory, Freiburg

CNFy

RhoA, activating

150 ng/ml

Aktories laboratory, Freiburg

FAK inhibitor 14

FAK

50 µM

Tocris, Bristol

Gliotoxin

Unknown

1 µM

Sigma-Aldrich, Taufkirchen

H-1152

ROCK

1 µM

EMD Millipore, Billerica

ML-141

CDC42 and Rac1 1 to 100 µM

NSC-23766

Rac1

5 to 500 µM

Tocris, Bristol

PF-562271

FAK

10 µM

Abcam, Cambridge

Q-VD-OPH

Caspases

20 µM

MP Biomedicals, Eschwege

SP600125

JNK

20 µM

Calbiochem (Merck), Darmstadt

Y-27632

ROCK

1 µM

EMD Millipore, Billerica

Santa Cruz, Heidelberg

57

Materials

5.4 Antibodies Antibodies were diluted in 3% milk/TBS supplemented with 0.1% Tween-20, if not stated otherwise.

Table 5.4 | Primary antibodies for Immunoblotting. Antigen

Supplier

Dilution

Species

Actin clone C4

MP Biomedicals

1:40000

Mouse

Bak-NT

Upstate

1:1000

rabbit

Bax-NT

Upstate

1:1000

rabbit

Bcl-2 10C4

Abcam

1:1000

mouse

Bcl-xL 54H6

Cell Signaling

1:1000

rabbit

Bim C34C5

Cell Signaling

1:1000

rabbit

Bim pT112/pS114

ProSci

1:2000 in BSA

rabbit

Cleaved Caspase 3

Cell Signaling

1:1000

rabbit

ERK1/2 137F5

Cell Signaling

1:1000

rabbit

ERK1/2 pT202/pY204

Cell Signaling

1:1000

rabbit

FAK D2R2E

Cell Signaling

1:1000

rabbit

FAK pTyr397 D20B1

Cell Signaling

1:1000

rabbit

Flag M2

Sigma

1:5000

mouse

GSK3α/β D75D3

Cell Signaling

1:1000

rabbit

GSK3β pS9 D85E12

Cell Signaling

1:1000

rabbit

HA-Tag C29F4

Cell Signaling

1:1000, in BSA

rabbit

JNK

Cell Signaling

1:1000

rabbit

JNK pT183/pY185

Cell Signaling

1:1000

rabbit

Mcl-1 D35A5

Cell Signaling

1:1000

rabbit

58

Materials MKK4

Cell Signaling

1:1000

rabbit

MKK4 pS257/pT261

Cell Signaling

1:1000

rabbit

MKK7

Cell Signaling

1:1000

rabbit

MKK7 pS271/pT275

Cell Signaling

1:1000

rabbit

MYPT1 D6C1

Cell Signaling

1:1000

rabbit

MYPT1 pT696

Cell Signaling

1:1000

rabbit

p190RhoGap

Novus Biologicals

1:1000

rabbit

p190RhoGap pY1105

Novus Biologicals

1:1000

rabbit

PARP

Cell Signaling

1:1000

rabbit

paxillin D9G12

Cell Signaling

1:1000

rabbit

paxillin pTyr118

Cell Signaling

1:1000

rabbit

RhoA 26C4

Santa Cruz

1:200

mouse

ROCK C8F7

Cell Signaling

1:1000

rabbit

Tubulin YL1/2

AbD Serotec

1:20000

rat

Peroxidase-conjugated secondary antibodies against mouse, rabbit or rat were used 1:5000 in 3% milk/TBS-T (0.1%) and were bought from Jackson ImmunoResearch.

59

Materials

5.5 Plasmids and shRNA vectors

Table 5.5 | List of plasmids Insert

Backbone

Use

Supplier

3 x HA Bim

pMIG

Retrovirus

Borner laboratory

Endo-14

mRuba

Endosomal

Addgene #55859

marker FAK

pCDH-EF1a-MCS-

Lentivirus

Gilmore laboratory

Borner laboratory

T2A-copGFP Flag-Bim

pCMV-2B

Expression

Flag-Bim T112A

pCMV-2B

Expression

Flag-Bim

pCMV-2B

Expression

pCMV-2B

Expression

GFP-paxillin

n.a.

Expression

Aktories laboratory

LifeAct

n.a.

Actin marker

Aktories laboratory

pM2DG

n.a.

Lentivirus

Borner laboratory

T112/S114A Flag-Bim S100/T112/S114A

envelope Packaging vector

psPAX

Lentivirus

Borner laboratory

production Packaging Vector

Hit60

Retrovirus

Borner laboratory

production VSVg

pCMV

Reroviral

Borner laboratory

envelope

60

Materials ShRNA against ROCK1 (SHCLNG-NM_005406.1) was purchased from SigmaAldrich (Taufkirchen, Germany). This shRNA is provided in the pLKO.1 backbone for lentiviral transduction. PLKO.1 control plasmid against luciferase was provided by U. Maurer (University of Freiburg). Sequence shROCK1: 5’CCGGGCACCAGTTGTACCCGATTTACTCGAGTAAATCGGGTACAACTGGTGCTTTTTG3’

5.6 Primers All QuickChange primers were designed using the Agilent QuickChange Primer Design Tool. Table 5.6 | Primers for side directed mutagenesis of Bim by QuickChange. Primer

Sequence 5’

3’

S100A fwd

TTT GAC ACA GAC AGG GCC CCG GCA CCC ATG AG

S100A rev

CTC ATG GGT GCC GGG GCC CTG TCT GTG TCA AA

T112A fwd

AGT TGT GAC AAG TCA ACA CAA GCC CCA AGT CCT C

T112A rev

GAG GAC TTG GGG CTT GTG TTG ACT TGT CAC AAC T

T112A/S114A fwd

CAA GTC AAC ACA AGC CCC AGC TCC TCC TTG CCA GG

T112A/S114A rev

CCT GGC AAG GAG GAG CTG GGG CTT GTG TTG ACT TG

S100E fwd

CTC TTT TGA CAC AGA CAG GGA GCC GGC ACC CAT GAG TTG TG

S100E rev

CAC AAC TCA TGG GTG CCG GCT CCC TGT CTG TGT CAA AAG AG

T112E fwd

GAG TTG TGA CAA GTC AAC ACA AGA GCC AAG TCC TCC TTG CCA GG

T112E rev

CCT GGC AAG GAG GAC TTG GCT CTT GTG TTG ACT TGT CAC AAC TC

61

Materials T112E/S112E fwd

CCA TGA GTT GTG ACA AGT CAA CAC AAG AGC CAG AGC CTC CTT GCC AGG CCT TCA ACC A

T112E/S112E rev

TGG TTG AAG GCC TGG CAA GGA GGC TCT GGC TCT TGT GTT GAC TTG TCA CAA CTC ATG G

Table 5.7 | Bim cloning primers Primer

Sequence 5’

3’

Bim-BamHI fwd

ATA TGG ATC CAA GCA ACC TTC TGA TGT AAG TTC TG

Bim-XhoI rev

ATA TCT CGA GTC AAT GCC TTC TCC ATA CCA GAC G

Traces of Aspergillus fumigatus DNA can be detected in peripheral blood by nested PCR, using the following primer pairs from Skladny et al.227.

Table 5.8 | nested PCR primer pairs for A.f. diagnosis. Primer

Sequence 5’

3’

AFU7 fwd

TTT GAC ACA GAC AGG GCC CCG GCA CCC ATG AG

AFU7 rev

CTC ATG GGT GCC GGG GCC CTG TCT GTG TCA AA

AFU5 fwd

AGT TGT GAC AAG TCA ACA CAA GCC CCA AGT CCT C

AFU5 rev

GGG AGT CGT TGC CAA CTC CCC TGA

62

Materials

5.7 Cell Lines All cell culture media are supplemented with 10% FCS, 60 µg/ml penicillin and 100 µg/ml streptomycin.

Table 5.9 | Adhesion cell lines Name

Cell type

AEC type 2

Murine

Medium Source

alveolar DMEM

Immortalization

Marco Idzko

primary

ATCC-CRL-9609

Viral

epithelial type 2 cells BEAS-2B

Human bronchial RPMI epithelial

HEK293T

Human

SV40

large

tumor antigen DMEM

Borner laboratory

Adenovirus type 5

Ulrich Maurer

Viral

embryonic kidney MEF Mcl-1-/-

Mouse C57BL/6 DMEM embryonic

SV40

large

tumor antigen

fibroblasts MEF Mcl-1-/- + Mouse C57BL/6 DMEM Mcl-1 wt

Ulrich Maurer

embryonic

Viral

SV40

large

tumor antigen

fibroblasts MEF MKK4-/-

MEF MKK4/7-/-

MEF MKK7-/-

Mouse C57BL/6 DMEM

Roger Davis

3T9:

9x105

cells

embryonic

seeded every third

fibroblasts

day

Mouse C57BL/6 DMEM

Roger Davis

3T9:

9x105

cells

embryonic

seeded every third

fibroblasts

day

Mouse C57BL/6 DMEM

Roger Davis

3T9:

9x105

cells

embryonic

seeded every third

fibroblasts

day

63

Materials MEF wild type

Mouse C57BL/6 DMEM

Roger Davis

3T9:

9x105

cells

embryonic

seeded every third

fibroblasts

day

Table 5.10 | Suspension cell lines Name

Cell type

Medium

Source

BAF3

IL-3 dependent, murine pro-B cell

DMEM + 1 ng/ml Ulrich Maurer IL-3

FL5.12

Jurkat

IL-3 dependent murine pro-B-cell DMEM + 1 ng/ml Ulrich Maurer lymphoid cell line

IL-3

Human T-cell lymphoma

DMEM

Ulrich Maurer

64

Methods

6 Methods 6.1 Cell Biology

6.1.1 Cell culture All adherent cells were cultured in TC flasks T175 with ventilated cap (Sarstedt, Nümbrecht) in the appropriate media (Table 5.9) at 37°C, 5% CO2 (HERAcell, Heraeus Instruments). Cells were split 1:10 before reaching confluency, usually every two to three days. For cell detachment, culture media was aspirated, cells washed in PBS and subsequently incubated with 1x trypsin-EDTA (2.2 mg/ml in PBS) at 37°C until visible detachment of cells, controlled by the Olympus CK30 light microscope (Olympus, Hamburg). Trypsinized cells were resuspended in 10 ml cell culture media to inhibit trypsin. This cell solution was either used to inoculate a new stock flask or counted and seeded in 10 cm culture dishes for further experiments. BEAS-2B bronchial epithelial cells can rapidly undergo squamous terminal differentiation when reaching confluency. Therefore, this cell line was only used until passage 20 to avoid any effects due to accidental differentiation. Suspension cells were cultured in 30 to 100 ml appropriate medium (Table 5.10) in TC flasks T175. All suspension cell lines were maintained at concentrations lower than 500,000 cells/ml. IL-3 dependent cell lines were transferred daily into fresh culture media. Therefore, cells were centrifuged for five min at 1200 rounds-per-minute (rpm; 300 x g). The supernatant was aspirated and cells were resuspended in an appropriate volume of media supplemented with fresh IL-3 (1 ng/ml). New cell lines were stored at liquid nitrogen to conserve early passages of cells. Therefore, cells were seeded on 15 cm culture dishes and grown to 70% confluency. After trypsinization, cells were washed in PBS and resuspended in 2 ml of the appropriate culture medium, supplemented with 20% FCS and 10% DMSO. 1 ml of this cell suspension was aliquoted in cryo tubes and frozen overnight at -80°C in an isopropanol chamber and subsequently transferred to liquid nitrogen (Biostor 5, Statebourne Cryogenics, Washington, UK).

65

Methods Cryopreserved cells were thawed in a 37°C water bath and instantaneously transferred to 10 ml media. After centrifugation, to remove DMSO, cells were resuspended in 10 ml media and seeded on 10 cm culture dishes.

1 x PBS: 1.47 mM KH2PO4, 2.68 mM KCl, 136.9 mM NaCl, 7.97 mM Na2HPO4

6.1.2 Cell death induced by Gliotoxin GT is unable to induce apoptosis, if cells are grown too dense. The seeded cell numbers as well as the confluency (max. 60 to 70%) on the day of the experiment have to be controlled to yield a consistent and reproducible cell death rate. In addition, cells were always seeded in the late evening before the day of the experiment. Usually, 1 x 106 cells were seeded in 10 cm culture dishes. Trypsin-detached cells (see previous chapter) were counted using the automated Z1 Coulter Counter (Beckman Coulter, Krefeld). The cell suspension was diluted 1:100 in 10 ml Coulter Isotone II Diluent. The setup of the machine is adjusted to count cells with a diameter between 9 and 27 µM. On the next morning, the cell media was aspirated and replaced by 5 ml fresh media for non-treated cells (NT) or 5 ml of media containing 1 µM Gliotoxin, if not stated otherwise. The GT treatment was started with an appropriate delay, in case of experiments with GT kinetics, and all time points were harvested together. Additional inhibitors and toxins were used in some experiments. In these cases the respective compounds were preincubated for one h before co-treatment of GT with the inhibitor. Inhibitors were prepared as 1000 x stock solutions (Table 5.3). The cytotoxic necrotizing factor y (CNFy) as well as the two component C3FT fusion toxin were a kind gift from Gudula Schmidt (University of Freiburg) and Carsten Schwan (AG Aktories, University of Freiburg). The Rho inhibiting C3FT toxin was prepared by combining 200 ng/ml C2II and 100 ng/ml C3FT in cell culture media. GT-treated cells detach rapidly. In order to assess the cell death rates, both the detached and the still attached cells were harvested. Therefore, the media with floating cells was collected in Falcon tubes. The remaining cells were washed in PBS, which was also collected in the same tube. Finally, cells were detached by trypsinization, if

66

Methods not stated differently, and transferred to the tube containing the floating cells. The harvested cells were centrifuged for 5 min at 1200 rpm and washed in PBS before they were used in further experiments. Cells were either subsequently assessed for cell death by flow cytometry or used to prepare protein extracts. Gliotoxin stock: 5 mg solved in 1.53 ml DMSO

10 mM, store at -80°C

6.1.3 Transfection of pro-apoptotic proteins Attractene was used to infect HEK 293T cells for transient protein expression according to the manufacturers protocol. Attractene is a nonliposomal lipid that facilitates the insertion of cDNA into adhesion cells. Attractene is suitable for sensitive cells because of its low toxicity. In brief, 4 µg vector was combined with 15 µl Attractene and adjusted to 300 µl volume using phenol-red free opti-MEM without serum. The mixture was added dropwise to 2 x 106 cells that were seeded the previous day. In case of overexpression of pro-apoptotic proteins like Bim, transfection was performed in the evening before the day of the experiment. In that case, Attractene transfection was not stopped after 4 to 6 h, but incubated overnight. In addition, induction of cell death by high levels of Bim was counteracted by the addition of QVD.

6.1.4 Transfection for confocal microscopy Lipofectamine was used to transfect BEAS-2B for confocal microscopy. Lipofectamine has a higher transfection efficiency than Attractene but can be toxic in some cases. For microscopy, 63,000 BEAS-2B cells were seeded on 2 cm glass bottom dishes (Greiner Bio One) and transfected with 6 µg of each plasmid on the next day. The appropriate volume of DNA was incubated with 5 ml opti-MEM; in a second tube 6 µl Lipofectamine was added to 250 µl opti-MEM. After 5 min incubation, both batches were combined and incubated for additional 15 min (don’t vortex). The mixture was added to the cells for 2 h instead of 4 to 6 h because BEAS-2B cells are sensitive to Lipofectamine. After washing and addition of 2 ml fresh medium cells were incubated overnight before microscopy. It is important to mention that Attractene cannot be used for confocal microscopy because its emission spectrum overlaps with some fluorescent dyes/proteins. 67

Methods

6.1.5 Lenti- and retroviral infection Two million HEK 293T cells were seeded for transfection to produce lentiviral or retroviral particles. Lentivirus was used for the delivery of PLKO.1 shRNA vectors and the overexpression of FAK. Attractene was used for transfection of HEK 293T cells as described above (chapter 6.1.3), with the difference that 3 µg of the plasmid of interest, 3 µg envelope vector pM2DG and 3 µg packing vector psPAX were mixed with 30 µl Attractene in 300 µl opti-MEM to produce lentiviral particles. Retrovirus was produced similarly by combining 3 µg of the plasmid of interest, 3 µg envelope vector VSVg and 3 µg packaging vector Hit60. On the next morning, protein biosynthesis was induced by adding 5 mM of the histone deacetylase (HDAC) inhibitor butyrate in 8 ml medium. Butyrate was removed after 8 h and replaced with 4 ml full medium. Virus particles were harvested the next morning by collecting the culture supernatant. Cell debris and other non-viral particles were removed by filtration with 0.45 µM SFCA syringe filters (Nalgene via Merck Millipore). The cleared supernatant was supplemented with 5 µg/ml polybrene. Polybrene is a cationic polymer that increases infection affinity by masking negative charges. This virus supernatant can either be stored at -80°C or preferentially be used directly for infection of target cells. For infection, cells were seeded the day before with a density of 12 x 105 cells/ml in six-well plates. In case of retroviral infection, the confluency should be less than 50% because retroviral vectors only integrate in the genome of proliferating cells. The cell culture medium was aspirated and replaced with 800 µl virus supernatant. The plate was centrifuged for 10 min at 2000 rpm and subsequently incubated for 1 h at 37°C. Afterwards 1 ml of medium was added and incubated overnight to allow the cells to divide. Cells can either be selected by the appropriate selection marker or be re-infected to yield higher expression levels. Cells were selected after one day with 4 µg/ml puromycin in case of shRNA mediated knock-down or were controlled by co-expression of GFP for overexpression of FAK.

68

Methods

6.1.6 Cell adhesion assay Human fibronectin (Stock = 500 µg/ml) was diluted 1:10 in PBS, aliquoted and stored at -20°C. For adhesion assays the fibronectin stock was further diluted to a final concentration of 10 µg/ml and used for the overnight coating of 96 well plates (37°C). On the next day, 50,000 suspension cells were seeded in fibronectin-coated 96-well plates and grown overnight. For quantification of the attachment, cells were washed with PBS and incubated with crystal violet solution for 30 min at room temperature. Residual dye was carefully removed with water. Completely dried cells were dissolved in 50 µl methanol for 30 min on a shaker and the absorbance was measured at 590 nm in the Tecan infinite M200 microplate reader (Tecan Group Ltd., Männerdorf). BAF3 suspension cells were attached to assess their response to GT. In that case, 100,000 cells were seeded in fibronectin-coated 12 well plates in presence of IL3. On the next day, cells were stressed with GT for up to 6 h and both floating as well as attached cells were harvested, pooling two wells per condition. The cells were then used for analysis by flow cytometry.

Cristal Violet staining Solution: 20% (v/v) Methanol, 0.5% (m/v) cristal violet.

6.1.7 Flow cytometry Fluorescence activated cell sorting (FACS) was used to quantify apoptotic cells by annexin-V-FITC staining. Treated and non-treated cells were harvested as described above (chapter 6.1.2). The resulting cell suspension was first washed in PBS and transferred to FACS tubes. Next, cells were washed in annexin-V binding buffer. The cell pellet was resuspended in 200 µl annexin-V binding buffer, containing annexin-VFITC (1:500) and incubated for at least 15 min in the dark. PI was added 1:10 from a 5 µg/ml stock, just before measurement. The percentage of apoptotic cells was then analyzed on a FacsCalibur or FacsLSRII (BD Bioscience, Heidelberg) and assessed by FlowJo software version 7.6.5 (FlowJo LLC, Ashland).

Annexin-V binding buffer: 2.5 mM CaCl2, 140 mM NaCl, 10 mM HEPES pH 7.4

69

Methods 6.1.7.1 Integrin surface staining BEAS-2B were treated with GT for up to six h and then harvested in PBS-EDTA (2.2 mg/ml) by scraping without trypsin digest. A possible loss of surface integrins by proteolytic digest is therefore reduced. The collected cells were passed through a 40 µm nylon cell strainer (BD Falcon, Heidelberg) to yield a single cell suspension. Cells were washed in FACS buffer and subsequently stained. For analysis of integrin expression on the cell surface, cells were incubated for 15 min in 100 µl FACS buffer with the following integrin antibodies.

Table 6.1 | Antibodies for integrin FACS staining Antigen

Tag

Supplier

Dilution

Species

Human CD62 clone VI-PL2

FITC

Bio Legend

1:100

Mouse

IgG ctrl

FITC

Bio Legend

1:100

Mouse

IgG ctrl

FITC

AbD Serotec

1:100

Hamster

Mouse CD29

FITC

AbD Serotec

1:100

Hamster

(Integrin β3)

(Integrin β1)

FACS buffer: 5 mM EDTA, 3% FCS.

6.1.7.2 Integrin activity assay Cyclo-RGD-5-FAM (Anaspec #65160) was used to assess integrin activity. The RGD peptide resembles the fibronectin binding motif of integrins. Active integrins, especially Integrin αV containing dimers, can bind to the fluorescently labelled peptide and therefore be quantified. 5-FAM is a commonly used green fluorescent compound. Treated cells were collected by scraping in PBS-EDTA and washed in FACS buffer. Cells were transferred into FACS tubes and subsequently stained with 5 µM

70

Methods RGD peptide in 200 µl FACS buffer for 15 min before analysis. The binding of RGD peptide to integrins is highly specific and the RGD peptide is too small to be detected in FACS, therefore the samples can be measured directly without an additional washing step. The 5-FAM fluorophore can be detected in the AlexaFluor 488 channel similar to FITC.

RGD stock: 1 mg dissolved in 510 µl H2O

2.5 mM.

6.1.8 Microscopy Brightfield micrographs were taken with the Nikon eclipse TS100 inverted microscope, equipped with the DS-L3 controller for image acquisition. Confocal time lapse microscopy has been performed in collaboration with Carsten Schwan (AG Aktories, University of Freiburg). BEAS-2B were transfected as described above (Chapter 6.1.4). Cells were analyzed 12 h post transfection. Confocal fluorescence microscopy was performed with an inverted microscope (Axiovert 200 M; Carl Zeiss) equipped with a spinning-disk head (Yokogawa) with emission filters, and solid-state laser lines (405, 488 and 561 nm). Fluorescence images were collected with a CoolSNAP-HQ2 digital camera (Roper Scientific) driven by VisiView imaging software (Visitron Systems). Glass Bottom dishes were incubated in a humidified atmosphere (6.5% CO2 and 9% O2) at 37°C. Images were processed with Metamorph software (Universal Imaging).

6.2 Biochemistry 6.2.1 Cell lysis and protein extraction Cell culture supernatant of floating cells and the attached cells were harvested (Chapter 6.1.2) and pelleted (1200 rpm, 5 min). Cell pellets were lysed on ice for 15 min in 50-100 µl whole cell lysis buffer. The whole-cell lysates were centrifuged for 10 min at 13,000 rpm. The supernatant was always kept on 4°C for further use on the same day. Protein concentration was determined using the Bradford reagent (Biorad).

71

Methods Table 6.2 | Whole cell lysis buffer Tris-HCL pH 7.5

20 mM

NaCl

150 mM

EDTA pH 8.0

5 mM

Na-pyrophosphate

5 mM

Na3VO4

1 mM

NaH2PO4

20 mM

β-glycerophosphate

3 mM

NaF

10 mM Add fresh before use:

Phosphatase inhibitor cocktail 1% (v/v) Complete protease inhibitor

1:25

MG-132

0.1% (v/v)

Triton-X-100

1% (v/v)

6.2.2 Bradford protein quantification After cell lysis, protein concentration was determined using the Bradford reagent (Biorad). The dye is deprotonated after binding to proteins, which results in a shift from red to blue (absorbance 595 nm). The increase of absorbance is proportional to the amount of protein in solution. Since this assay yields relative values, a dilution series of bovine serum albumin (BSA) was used from 0 to 80 µg/ml to estimate the unknown sample concertation. 100 µl of the standard or diluted samples (1:250 – 1:1000) were mixed with 100 µl Bradford reagent (1:2.5 pre-diluted in water) and measured using the SpectraMax 340 PC microplate spectrophotometer (Molecular Devices, Biberach an der Riß).

72

Methods

6.2.3 Gel electrophoresis & Immunoblotting Protein lysates were separated by polyacrylamide gel electrophoresis (PAGE). This method allows the separation of proteins by their molecular weight, owing to the reducing conditions and neutralization of positive charges by sodium dodecyl sulfate (SDS). Equal amounts of protein lysate (40-60 µg) were adjusted to the same volume and were boiled with 1 x SDS sample buffer at 97°C for 5 min prior to separation by SDS-PAGE at 140 V. The separation gel contains 10% to 15% acrylamide, depending on the size of the protein that is to be detected. The stacking gel contains 4% acrylamide. Table 6.3 | Gel compositions for different acrylamide concentrations 15%

12%

10%

stacking

Solution A (B for stacking)

4.17 ml 4.17 ml

4.17 ml

4.17 ml

30% acrylamide

8.25 ml 6.4 ml

5.5 ml

1.7 ml

H2O

3.16 ml 5 ml

5.9 ml

5.1 ml

10% SDS

166 µl

166 µl

166 µl

100 µl

15% APS

83.5 µl

83.5 µl

83.5 µl

50 µl

10% Temed

100 µl

100 µl

100 µl

100 µl

Solution A: 1.5 mM Tris-HCL pH 8.9, 8 mM EDTA. Solution B: 0.5 mM Tris-HCL pH 6.8, 8 mM EDTA. 1 x running buffer: 25 mM Tris, 190 mM Glycine, 0.1% SDS. 3 x SDS sample buffer: 10 mM Tris, 1 mM EDTA, 3% SDS, 13% glycerol, 0.075% bromphenol blue, 7 mg/ml DTT.

Proteins smaller than 15 to 20 kD were separated by Tricine-SDS-PAGE as described by Hermann Schägger228. The separated proteins were either directly stained by Coomassie or were transferred in 1 x transfer buffer at 400 mA by wet tank blotting on 0.2 µm nitrocellulose

73

Methods membranes (GE Healthcare) for 1 h. The blots were blocked for 1 h in blocking buffer and subsequently incubated overnight at 4°C with the primary antibody of choice (Table 5.4) Peroxidase-conjugated secondary antibodies against mouse, rabbit or rat were used for immunodetection using Enhanced Chemiluminescence (Pierce, Rockford, IL, USA), X-Ray films (Fuji Medical) and the Curix60 (AGFA) for development. In case of quantification, blots were developed on FusionSL Vilber Lourmat (PeqLab) and quantified using the software FusionCapt Advance Solo 4 (V.16.08). 1 x transfer buffer: 25 mM tris-HCL, 190 mM Glycine, 20% Methanol Blocking buffer: 1 x TBS, 0.1% Tween-20, 5% milk.

6.2.3.1 Coomassie and silver staining Protein samples were separated by SDS-PAGE and subsequently fixed in 40% ethanol and 10% acetic acid for 1 h. After 2 washes for 10 min in water, gels were stained with Coomassie staining solution overnight. Gels were destained in water until the background completely disappeared. Silver staining was performed using the SilverQuest SilverStaining Kit according to the manufacturer’s protocol with the modification that only one quarter of all volumes was used.

Coomassie staining solution: 10% Acetic acid, 0.025% Coomassie blue R250 (m/v), 40% Methanol

6.2.4 Immunoprecipitation Whole cell lysates, containing a minimum of 300 µg protein, were used for immunoprecipitation (IP). 10% of the lysate was used as input control for immunoblot. Agarose beads (50 µl) and appropriate antibodies were used to IP untagged proteins. The cell lysate was adjusted to 300-500 µl and pre-cleared for 1 h in bead slurry without antibody (50 µl slurry). The supernatant was then used for a 30 min incubation with 2 µg antibody at 4°C. Beads were added (50 µl) and incubated for 2 h at 4°C. Finally,

74

Methods beads were washed three times in cell lysis buffer (8,200 x g, 1 min) and boiled in SDS sample buffer for immunoblot. FLAG-tagged proteins were enriched using 40 µl EZ view red anti-FLAG Affinity gel (Sigma) according to the manufacturer’s protocol. Proteins were eluted by incubation with 50 µl 3 x FLAG peptide. Whole cell lysis buffer for pulldown of FLAGBim was prepared using 1% CHAPS instead of Triton-X-100. HA-tagged proteins were purified from 300 µl whole cell lysate using anti-HA affinity matrix (Roche) as stated in the manufacturer’s protocol.

6.2.5 Rhotekin pulldown Rhotekin-RBD Protein GST beads (Cytoskeleton Inc., Denver, CO, USA) were used to assess Rho activity in whole cell lysates. Rhotekin-RBD specifically binds to activated, GTP-bound RhoA and RhoC and to lower extent to RhoB but not to Rac or CDC42229. Rhotekin-coupled beads thereby allow to enrich the activated population of Rho. Samples were adjusted to the same cell numbers after harvesting the treated cells. Bradford quantification was omitted to perform the pulldown without delay. The cells were lysed in 300 µl GST-Fish buffer for 5 min and centrifuged at 13,400 rpm for 1 min. 20 µl of this lysate was used for input loading control. The remaining supernatant was supplemented with 60 µg pre-washed Rhotekin beads (19 µl beads per sample, resuspended in 100 µl GST-Fish buffer) and incubated for 45 min at 4°C on an over-head shaker. After washing once in GST-Fish buffer, the samples were boiled in 12 µl SDS-sample buffer and RhoA was visualized on immunoblots.

75

Methods Table 6.4 | GST-Fish buffer Glycerol

10% (v/v)

NaCl

100 mM

NP-40

1% (v/v)

MgCl2

2 mM

Tris-HCl pH 7.4

50 mM

Add fresh PMSF

1 mM

Rhotekin beads: 2 mg lyophilized beads in 600 µl H2O

3.3 µg/µl stock, store in

40 µl aliquots at -80°C.

6.2.6 Mass spectrometry Mass spectrometry (MS) was performed to analyze the covalent modification of integrins by GT. Human recombinant αVβ3 integrin (5 µg) was mixed with 5 µg BSA as an internal control in 50 mM Hepes, pH 7.5. First, thiols were reduced by the addition of 10 mM DTT at room temperature for 1 h. GT is a redox sensitive compound and isinactivated after reduction of the reactive disulfide bridge. Therefore, DTT was removed by filtration on VivaSpin 500 columns with a cut-off of 30 kD (Sartorius Biotech GmbH, Göttingen). The restrained recombinant proteins were washed eight times in 50 mM HEPES pH 7.5 and subsequently collected in 100 µl HEPES. Proteins were denatured in 0.1% RapiGest (Waters, Eschborn) for 45 min at 70°C. Peptides were generated by digestion with 0.2 µg trypsin overnight at 37°C. Samples were acidified to pH 3 using 3 M Guanidinium-HCL for RapiGest degradation, followed by purification of the peptides using C18 columns. The purified peptides were dried in a vacuum centrifuge and resuspended in 100 µl HEPES pH 8 (100 mM). Finally, peptides were incubated with 100 to 1,000 µM GT overnight at 37°C to modify free thiols of cysteine containing peptides. Finally, the GT-treated peptides were purified using C18 column prior to MS. The Q-Exactive plus system (Thermo Scientific, Bremen, Germany) was used for mass spectrometry in collaboration with Maria Magdalena Koczorowska (AG

76

Methods Schilling). The mass spectrometer was operated in the data-dependent mode as described previously230. Peptide sequences were identified by X! Tandem (Version 2013.09.01)231 using the reviewed canonical human and bovine combined Uniprot sequences (without isoforms) together with an equal number of randomized decoy sequences, generated by DBtoolkit232. Identified peptides were mapped on the protein sequence using the software Proteator (unpublished, AG Schilling)233.

6.2.7 Caspase activity assay To assay for caspase activity, 20-30 µg protein lysate from whole cell lysates was adjusted to a volume of 10 µl in caspase buffer. Ac-DEVD-AMC fluorogenic caspase3/-7 substrate (6 µM final, Enzo Life Sciences) was added in 90 µl caspase activity buffer and immediately read in black, flat bottom 96 well plates using the Tecan infinite M200 microplate reader. Increase of fluorescence intensity was measured every second minute for a total of 30 min. The slope of the linear increase was calculated and displayed relative to the sample with highest caspase activity (= 100%). The DEVD sequence resembles the PARP cleavage site of Caspase-3 but can also be cleaved by caspase-1/-4/-7 and -8. The fluorogenic tetra peptide Ac-DEVD-AMC has an emission maximum at 440 – 460 nm after cleavage.

Caspase activity buffer: 100 mM HEPES pH 7.5, 10 mM DTT (add fresh).

6.2.8 JNK and ROCK activity assays The KinaseSTAR JNK Activity Assay Kit was used to measure JNK activity in response to GT treatment. The kit was used as stated in the manufacturer’s protocol. In brief, JNK is enriched by pulldown from whole cell lysates and incubated with recombinant cJun. Phosphorylation of cJun was quantified by Western blot. The activity of ROCK kinase was determined by detecting the phosphorylated form of its substrate myosin-binding subunit of myosin phosphatase 1 (MYPT1, pT696) on immunoblots of whole cell lysates and quantified relative to total MYPT1.

77

Methods

6.2.9 Radioactive phosphorylation assay FLAG-tagged Bim wild-type and phosphorylation deficient mutants were enriched by FLAG pulldown from HEK293T cells, transfected by Attractene in the presence of caspase inhibitor QVD (chapter 6.1.3.). Cell lysates were prepared in whole cell lysis buffer without EDTA to prevent complexation of cations. After FLAG IP, Bim was eluted in 75 µl 3 x FLAG peptide, 15 µl of this elution was used in the kinase assay (+/kinase). The same volume of kinase reaction mix was added to the eluate and incubated for 20 min at 30°C prior to immunoblotting. The dried membrane was incubated overnight at -80°C and the phosphorylation signal was quantified using the PhosphorImager (PerkinElmer, Waltham). Subsequently, membranes were blocked and incubated with primary antibodies to detect the protein of interest. Table 6.5 | 10 x kinase buffer DTT

12 mM

Glycerol

10% (v/v)

HEPES pH 7.5

600 mM

MgCl2

30 mM

MnCl2

30 mM

NaVO4

30 µM Adjust to pH 7.5

Kinase reaction mix: 2 x kinase buffer, 200 µM ATP, 12.5 µCi γ32pATP (#SRP-301), +/- 100 ng JNK2.

78

Methods

6.3 Molecular Biology 6.3.1 Plasmid preparation Plasmids were either prepared from glycerol stocks or by transformation of competent E.coli with plasmid DNA. Non-viral vectors were amplified using OneShot Top10 competent cells (Invitrogen) and viral vectors were transformed in MAX Efficiency Stbl2 competent cells (Invitrogen) to avoid recombination of the plasmid. All cells were transformed by heat-shock for 30 sec at 42°C after 30 min pre-incubation on ice. The transformation mixture was added to 500 µl pre-warmed SOC medium, incubated for 1 h and plated on agar plates containing the appropriate antibiotic. Up to 10 different clones were cultured from the agar plate on the next day in 2 ml LB broth for purification of the plasmid DNA using the Wizard Plus Miniprep DNA Purification Kit. All plasmids were verified by sequencing, performed by GATC Biotech (Mulhouse). Finally the 2 ml culture broth was used to inoculate 250 ml overnight cultures for purification of the correct plasmid via the ZymoPure Maxiprep Kit.

6.3.2 Side directed mutagenesis The Quick Change Site directed mutagenesis Kit was used to introduce point mutations in the gene encoding for Bim. The mutagenesis primers for phosphorylation deficient mutations (S/T

A) as well as phosphorylation mimicking mutations (S/T

E) are

listed in Table 5.6. The side directed mutagenesis was performed as stated in the manufacturer’s protocol using the following PCR thermo-profile.

Table 6.6 | QuickChange PCR profile

18 x

95°C

1 min

95°C

50 sec

60°C

50 sec

68°C

7 min

68°C

7 min

79

Methods The template DNA was digested after PCR by DpnI digest for 1 h at 37°C prior to heatshock transformation of XL10-Gold or STBL2 bacteria for viral vectors. A minimum of five clones were cultured for plasmid preparation from agar plates and were subsequently sequenced.

6.3.3 Cloning Bim and the different Bim mutants were subcloned into various vectors e.g. for retroviral infection or to yield tagged proteins. The cloning primers are shown in Table 5.7. The following PCR reaction was prepared to amplify Bim from template DNA.

Table 6.7 | cloning PCR reaction mix 10x Pfx Amplification Buffer 1 x dNTP mix

0.3 mM

MgSO4

1 mM

Forward primer

0.3 nM

Reverse primer

0.3 nM

Template DNA

50 ng

Platinum Pfx DNA Pol

0.4 µl

Add H2O to 50 µl

Table 6.8 | cloning PCR profile

30 x

95°C

5 min

95°C

15 sec

55°C

30 sec

68°C

1 min

68°C

10 min

80

Methods The resulting PCR product was verified by agarose gel and extracted from the gel using the Gel extraction Kit (Qiagen) and subsequently digested for 15 min at 37°C . Table 6.9 | Restriction digest Restriction buffer

1x

BamHI

1 µl

XhoI

1 µl

Vector/PCR product

1 µg/2 µl

Add H2O to 20 µl

The cut vector and PCR product were purified by agarose gel and isolated with the gel extraction kit for ligation. The ligation was performed using a 3:1 ratio of insert to vector. Vector and insert were ligated by T4 Ligase for 5 min at room temperature. Finally the ligation product was transformed into appropriate competent bacteria.

6.4 In vivo experiments 6.4.1 Fungus cultivation Aspergillus fumigatus strain B-5233 was cultured in TC flasks T175 with ventilated cap, filled with Sabouraud agar at 37° for three days until a green solid biofilm is formed (Figure 6.1). Cultivation of A.f. was performed in a S2 certified fume hood (AG Koch). Conidia can be harvested with a bacteria inoculation loop to maintain the culture. Alternatively, the surface can be rinsed with 10 ml PBS containing 0.05% Tween-20 to facilitate re-suspension of the hydrophobic spores. The conidia were then counted in a Neubauer chamber. The conidia solution was either used to inoculate a liquid culture for GT production or for the infection of C57BL/6 mice. GT was produced by inoculation of 50 ml RPMI, supplemented with 25 mM HEPES, with 1 x 107 conidia. The conditioned medium was harvested after 72 h and

81

Methods purified from fungal hyphae by sterile filtration. FCS and P/S were added for the use of conditioned medium in cell culture experiments.

Figure 6.1 | A.f. B-5233. A.f. was cultured for 3 days on Sabouraud agar at 37°C.

6.4.2 Invasive Aspergillosis model in C57BL/6 mice All experiments involving primary cultures and animals were performed with the approval of the “Regierungspräsidium Freiburg” and were conform to the animal experimentation laws. The C57BL/6 mouse strain was used during these studies in order to employ existing knock-out mouse strains, although other mouse strains like the 129/Sv strain were reported to be more susceptible to develop IA. Mice were immunosuppressed for four days prior to and four days post infection. In this context the day 0 refers to the day of infection with A.fumigatus. Each animal received a subcutaneous (s.c.) and an intraperitoneal (i.p.) injection with 3 mg hydrocortisone-acetate (HC) in 100 µl PBS containing 0.05% Tween-20 per injection. In total, each mouse received 6 mg HC at day -4, -2, 0, 2 and 4 (Figure 6.2).

82

Methods

Figure 6.2 | In vivo infection scheme. C57BL/6 mice were immunosuppressed four days prior and four days post-infection with 6 mg HC on day -4, -2, 0, 2 and 4. Animals were infected with 6.5 x 108 A.f. conidia on day 0. Immunosuppression was controlled by taking peripheral blood (PB) on day 0 before infection and on day six to monitor recovery of the animals. All animals were checked four times a day to assess disease progression. Animals were sacrificed if cancellation criteria were met or after 30 days post-infection to collect PB, bronchial alveolar lavage fluid and the lung for cryosections.

Each mouse was infected on day 0 with 6.5 x 108 conidia, which were harvested on the same day in PBS/0.05% Tween-20. Control mice were immunosuppressed but treated with PBS/T only to control possible effects of the HC injection. All animals were infected using 15 ml glass bottles with homogenizer. The health of all animals was monitored four times per day and mice were sacrificed according to cancellation criteria which define the disease free survival. Mice were sacrificed by i.p. injection of 500 µl thiopental (200 mg/kg body weight). Peripheral blood (PB) was collected before thiopental injection from the orbital sinus to test for A.f. infection. Thiopental solution: 0.5 g solved in 63 ml 1 x PBS, 0.5 ml = 5 mg per mouse

83

Methods

6.4.3 White blood cell counting Peripheral blood (PB) was used to control the health status of immunosuppressed animals. PB was collected in Microvette CB300 Li-Heparin tubes (Sarstedt, Nümbrecht). Samples were diluted 1:4 in 1 x PBS in a total volume of 200 µl and analyzed using the ADVIA 120 hematology system (Siemens, Berlin) and the setup “mouse500 CBC/Diff/Retic”. This setup quantifies the amount of neutrophils, monocytes, basophils, eosinophils and lymphocytes and thereby reflects the immune status of the animals.

6.4.4 Broncho alveolar lavage and lung preparation Thiopental sacrificed mice were used to harvest broncho alveolar lavage fluid (BALf). The trachea was perforated and a vein flow (22G, 36 ml/min, Braun, Melsungen) was carefully inserted into the trachea without reaching the pulmonary airways. The vein flow was hold in place with an operation suture. 1 ml of BAL solution was injected at once with steady pressure into the lung and aspirated. Next, the thorax was opened for preparation of the lung. The lung was inflated by injection of 1 to 2 ml O.C.T. Tissue TEK compound (Sakura, Staufen im Breisgau). The trachea was closed using the operation suture and the lung was immediately transferred to Aluminum Cryo-Tubes (Hycultec, Beutelsbach) on dry ice. BAL solution: 1 x PBS, 0.5 mM EDTA

6.4.5 Detection of A.f. infection A.f. can either be diagnosed by a nested PCR, amplifying fungal DNA, or by Platelia ELISA against a fungal galactomannan antigen. The ELISA assay is more reliable and used in the clinics for IA diagnostics. The assay can either be performed using PB or BAL. In each case, 300 µl of sample fluid were used in the assay, which was carried out exactly as stated in the manufacturer’s protocol. The absorbance of the sample was quantified relative to, in the kit provided, positive and cut-off controls. Using the BAL and PB of sacrificed mice allowed the detection of A.f. in the lung and permitted the distinction if inhaled candida became invasive or not.

84

Methods

6.4.6 Histological sections and fungus staining The lungs of sacrificed mice were prepared for histological sections to assess the invasive potential of the infection. Lungs were embedded and frozen in tissue freezing media (Leica, Wetzlar) and cut in 10 µm slides using the cryotome (Leica, Wetzlar). Three to four lung slides were transferred on one SuperFrost plus microscopy slide (R.Langenbrick, Emmendingen). The Accustain Silver Stain kit was used to visualize fungal hyphae, spores and other opportunistic organisms. The staining was performed as stated in the manufacturer’s protocol with the modification that slides were stained for 40 min in the silver methenamine solution at 62°C instead of 20 min. In addition, the slides were stained with haematoxylin (3 min) and eosin (1 minute). Finally the slides were embedded in Entelan (Merck, Darmstadt) and covered with precision coverslips LH26.1 (Carl Roth, Karlsruhe). The slides were analyzed on an Axioscop 2 MOT (Zeiss, Jena), equipped with AxioVison Rel. Software (version 4.8).

85

Bim is released from pro-survival Bcl-2

Results 7 Bim is released from pro-survival Bcl-2 GT-induced apoptosis requires the activation of the pro-apoptotic proteins Bim and Bak132,226. Bim can execute apoptotic signals by a variety of possible mechanisms, which are all based on the interaction of Bim with pro-survival proteins of the Bcl-2 family, as well as direct activation of the effector proteins Bax and Bak (chapter 2.2.1). The mode of cell death regulation is highly dependent on the cell type and the stimulus and has thus to be investigated individually in every setup. Here we thought to identify the differential Bcl-2 family interaction profile of Bim in response to Gliotoxin (GT). Therefore, HEK293T cells were transiently transfected with wild-type (wt) FLAG-tagged Bim. At 12 h post transfection, the cells were treated with 1 µM GT for indicated time points. Bim was enriched by anti-FLAG pulldown and interacting proteins were analyzed on immunoblots (Figure 7.1).

Figure 7.1 | GT releases Bim from Bcl-2. FLAG-tagged Bim wild-type was enriched by FLAG IP from transfected HEK293T after indicated time points. Cells were stressed with 1 µM GT prior to IP. Pro-survival Bcl-2 interacted with Bim in non-treated cells (NT) and was released in the early response to GT. However, Bcl-2 was recruited back to Bim at later time points. Pro-apoptotic Bak, in contrast, showed increased binding to Bim during early time points of the kinetic. Non-transfected and non-treated cells served as a control for the specificity of the FLAG-IP. 10% of the whole cell lysate was used as input to control protein levels (right panel).

86

JNK phosphorylates Bim to execute apoptosis While Bcl-2 effectively bound Bim in healthy cells, a substantial amount of Bcl-2 was released from immunoprecipitated Bim within the first hours of GT treatment and was recruited back to Bim at later time points. In contrast, Bak gradually bound more FLAG-Bim during the first 3 h of GT treatment. These data indicated that in response to GT, Bim is relieved from its inhibitory interaction with Bcl-2 which most likely provides Bim the capacity to better interact with and activate Bak.

8 JNK phosphorylates Bim to execute apoptosis Parts of the data shown in this chapter were published during the course of the thesis132. They indicated that JNK transmits the pro-apoptotic signaling of GT by phosphorylating Bim at S100, T112 and S114 in human bronchial epithelial cells (BEAS) and MEFs. We found that both isoforms of JNK, JNK1 and JNK2 were consistently phosphorylated/activated in response to GT. Thus, for the sake of simplicity we will therefore describe JNK1/2 as “JNK” phosphorylation/activation. Radioactive γ32ATP was employed in kinase assays to verify these phosphorylation sites in vitro. FLAG-tagged wt Bim and phosphorylation-deficient mutants were pulled down with anti-FLAG beads from HEK293T cells and incubated with the recombinant kinases ERK or JNK. ERK was reported to phosphorylate Bim at S55, 65 and 73131. Indeed, ERK effectively phosphorylated Bim but not the JNK-specific target ATF2 (Figure 8.1). FLAG-tagged Bim was phosphorylated by JNK as mutation of the putative JNK phosphorylation sites decreased the γ32ATP phospho-signal (Figure 8.1). If all three sites (S100, T112 and S114) were mutated to alanines, the JNK-mediated phosphorylation signal was totally lost. Bim was detected on the same membrane to control for equal protein levels. In vitro phosphorylation of Bim by JNK indicated that this pathway might be relevant in cells as well. Therefore, primary murine alveolar type II cells (AEC type II) were used to study the involvement of JNK on GT signaling in non-transformed, physiologically relevant cells.

87

JNK phosphorylates Bim to execute apoptosis

Figure 8.1 | JNK phosphorylates Bim in vitro. FLAG-Bim and phosphorylation deficient mutants were enriched by anti-FLAG IP from HEK293T cells and incubated with recombinant kinases in the presence of γ32ATP. ERK is known to phosphorylate Bim (positive control) but failed to phosphorylate the JNK-specific effector ATF2. Recombinant ATF2 served as a negative control for ERK activity. Wild-type Bim was phosphorylated by JNK. The γ32ATP signal decreased with an increasing degree of mutation and was abrogated in the triple mutant. JNK-Bim signaling was analyzed by immunoblotting lysates from GT-treated (5 µM) AEC type II cells. JNK was phosphorylated within 2 h after the treatment, detected by an antibody directed against phosphorylated Thr183 and Tyr185 residues (Figure 8.2). Co-treatment of cells with GT and the JNK inhibitor SP600125 (100 µM) abolished phosphorylation of JNK as well as the phosphorylation of Bim. Bim phosphorylation was monitored using a custom-made phosphorylation specific antibody against the pS112/pT114 site (row 4, Figure 8.2) and by gel mobility shift analysis (row 3, Figure 8.2).

88

JNK phosphorylates Bim to execute apoptosis

Figure 8.2 | JNK activates Bim. AEC type II cells were treated with 5 µM GT for the indicated time points. Immunoblotting showed phosphorylation/activation of JNK (Thr183/Tyr185) as well as Bim (T112/S114) concomitant with the processing and hence activation of caspase-3. Bim phosphorylation was assessed by a phosphorylation-specific antibody (row 4) and detecting a shift of Bim protein bands to higher molecular weights on the SDS gel (row 3). Co-treatment with the JNK inhibitor SP600125 (SP, 100 µM) abolished phosphorylation of JNK and Bim and decreased the processing of caspase-3. In parallel to JNK and Bim phosphorylation, we observed the processing of caspase-3 from its 32 kD pro-form to the active 17 kD form. This caspase-3 activation was decreased if GT was combined with the JNK inhibitor SP600125, indicating that GT-induced caspase-3 processing depended on JNK activation. Caspase-3 cleavage, as assessed by immunoblotting, correlated with increased caspase-3 activity. GT-treated AEC type II lysates were employed to measure caspase-3 activity, using the fluorogenic caspase substrate DEVD (Figure 8.3). The highest caspase activity was detected after 6 h of treatment with 5 µM GT. Caspase activity was decreased if GT was combined with SP600125 (100 µM), which was in line with decreased processing of caspase-3 under these conditions (Figure 8.2).

89

JNK phosphorylates Bim to execute apoptosis

Figure 8.3 | JNK inhibition protects AEC type II cells from GT-induced apoptosis. Caspase activity was assessed by the fluorogenic DEVD peptide in lysates from AEC type II cells treated with GT for 6 h. The combination of GT with the JNK inhibitor SP600125 decreased the activation of caspases significantly. Graphs show the means of at least three independent experiments ± S.E.M; p-values: *0.05-0.01, ** 0.01-0.001, *** < 0.001, two-way anova, posthoc: Bonferroni compared to untreated cells. The protective effect of JNK inhibition on caspase processing and caspase activity can also be seen morphologically. AEC cells treated with GT detached within the first hour of treatment and showed apoptotic membrane blebs after 4 h of exposure to GT (Figure 8.4). Cells still detached when they were co-treated with SP600125 (20 µM) but they did not show morphological features of apoptosis (Figure 8.4, C and D). These data suggest that JNK was activated by GT leading to the triple phosphorylation of Bim in order to transduce the apoptotic stimulus to Bax/Bak activation. We therefore asked the question how the toxin is able to activate this stress response protein kinase.

90

Unraveling the signaling events leading to GT-induced JNK activation

Figure 8.4 | JNK inhibition protects from GT-induced formation of membrane blebs, but not from cell detachment. (A) AEC type II cells were treated with 1 µM GT for 4 h. All cells detached within the first hour after addition of GT and showed morphological changes associated with apoptosis after detachment (B). (C) And (D) co-treatment with JNK inhibitor SP600125 (20 µM) protected the detached cells from forming apoptotic membrane blebs in detached cells. Scale bars represent 100 µm.

9 Unraveling the signaling events leading to GT-induced JNK activation 9.1 JNK is phosphorylated by MKK4 and MKK7 Immunoblotting of AEC cell lysates revealed that JNK was phosphorylated after treatment with GT. Therefore we sought to identify the kinase(s) that mediate activation of JNK. JNK is a stress response kinase that belongs to the family of MAPKs. MKK4 and MKK7 were reported as MAPKKs for JNK in response to different stimuli (chapter 2.3.1). Wild-type MEF and MEF deficient for these MAPKKs were employed to test this hypothesis (Figure 9.1). Cells deficient for MKK4, MKK7 or both were treated for 6 h with 1 µM GT. Both wild-type and knock-out MEF detached within the first hours of GT

91

Unraveling the signaling events leading to GT-induced JNK activation treatment with a similar kinetic as AEC type II cells. However, only the wild-type MEFs showed apoptotic membrane blebs after 6 h (Figure 9.2).

Figure 9.1 | Verification that MEFs were deficient for MKK4, MKK7 or both. Immunoblots of untreated MEF lines showed that cells were deficient for either MKK4, MKK7 or both. These cells were used in the following experiments.

Figure 9.2 | MKK4/7 deficient MEF detach in response to GT. Different MEF cells were treated with 1 µM GT for 6 h. Wild-type cells detached in response to GT and showed an apoptotic morphology. All knock-out MEF detached similarly to wild-type MEF but did not show apoptotic membrane blebs. In addition to the morphological changes, 71% of wt MEF were positive for annexin-V staining and showed a high degree of caspase activity (Figure 9.3). MEF cells lacking either MKK4 or MKK7 did not exert major annexin-V and caspase activities (Figure 9.3), indicating that they were protected from GT-induced apoptosis.

92

Unraveling the signaling events leading to GT-induced JNK activation The highest degree of cell death protection was seen with MEF deficient for both MKK4 and MKK7. We therefore think that MKK4 and MKK7 are at least partially redundant in mediating GT-induced apoptosis signaling.

Figure 9.3 | MKK4/7 deficient MEF are protected from GT-induced caspase activation and apoptosis. MEF deficient for MKK4, MKK7 or both were treated with 1 µM GT for 6 h and subjected to annexin-V staining or caspase activity assay. (A) Single or combined deletion of these MAPKKs for JNK reduced the annexin-V positive cell population. (B) Additionally, these cells showed a significantly reduced activation of caspases. The combined loss of MKK4 and MKK7 rescued cells most efficiently from GT-induced apoptosis. Graphs show the means of at least three independent experiments ± S.E.M; p-values: *0.05-0.01, ** 0.01-0.001, *** < 0.001, two-way anova, posthoc: Bonferroni compared to wild-type (WT). Immunoblotting using phosphospecific antibodies against the active forms of MKK4 and MKK7 revealed that both MKK4 and MKK7 were phosphorylated and hence activated after treatment of human bronchial epithelial cells with GT. MKK4 showed phosphorylation at Ser257/Thr261 and MKK7 showed phosphorylation at S271/T275 after 4-6 h of GT treatment (Figure 9.4, A). In parallel, we detected pro-apoptotic PARP processing, indicating that MKK4/7 phosphorylation was responsible for caspase-3 activation. To further prove the importance of MKK4/7 in JNK-Bim signaling, we monitored GT-induced JNK and Bim phosphorylation in the MKK4/7 knock-out cells, as shown in Figure 9.4, B. While wt MEFs exhibited strong JNK and Bim phosphorylation

93

Unraveling the signaling events leading to GT-induced JNK activation concomitant with caspase-3 activation and PARP cleavage, all processes were completely ablated in MEFs deficient for both MKK4 and MKK7.

Figure 9.4 | MKK4 and 7 are phosphorylated and act upstream of JNK. Immunoblot analysis of cells treated with 1 µM GT for 4-6 h. (A) BEAS-2B cells showed phosphorylation of MKK4 and MKK7 in response to GT. Phosphorylation of MKK4 was detected by an antibody directed against pS257/pT261 and MKK7 phosphorylation was detected using a MKK7 pS271/pT275 specific antibody. Phosphorylation of MKKs correlated with apoptosis-associated cleavage of PARP. (B) Knock-out of MKK4 and MKK7 abolished the activating phosphorylation of JNK (Thr183/Tyr185) in MEFs. Similarly, phosphorylation and gel mobility shifts of Bim were highly reduced (lower panel). Absence of caspase-3 processing and PARP cleavage in these cells underlined the protective effect of MKK depletion compared to wild-type (wt) MEFs. These data revealed that MKK4 and MKK7 co-operated to transmit the GT-induced apoptotic signaling to JNK. Since both MKKs were phosphorylated after treatment with GT, there must be another protein kinase which acts upstream of these MAPKKs to phosphorylate and activate them.

94

Unraveling the signaling events leading to GT-induced JNK activation

9.2 Inhibition of GSK3 protects from GT-induced apoptosis The glycogen synthase kinase-3 (GSK3) is regulated downstream of AKT and PI3K in response to growth factor deprivation. GSK3 mainly induces cell death via the induction of Puma and the degradation of the survival factor Mcl-1 (chapter 2.3.1). Interestingly, GSK3 co-operates with JNK in order to phosphorylate Mcl-1, thereby targeting Mcl-1 for proteasomal degradation146. We therefore tested the possibility that GSK3 was involved in the JNK-dependent pro-apoptotic signaling of GT. Inhibition of GSK3 by the specific inhibitor Axon-1126 (0.75 µM) rescued GT-induced cell death. BEAS-2B detached after treatment with GT and displayed membrane blebs after 6 h. Cells co-treated with GT and Axon-1126 still detached but did not show apoptotic morphological changes (Figure 9.5, A). In line with these data caspase activation was blocked when GT was combined with the GSK3 inhibitor (Figure 9.5, B).

Figure 9.5 | Inhibition of GSK3 protects from GT-induced apoptosis. BEAS-2B cells were treated with 1 µM GT for the indicated time points. (A) Morphological changes like membrane blebbing, indicative for apoptosis, were blocked by the GSK3 inhibitor Axon-1126 (0.75 µM). (B) Axon-1126 also blocked GT-induced activation of caspases. Scale bars = 200 µm. Graphs show the means of five independent experiments ± S.E.M; p-values: *** < 0.001, student’s t-test compared to GT + Axon. When testing the impact of GSK3 inhibition on GT-induced JNK/Bim phosphorylation/activation, we found that it did not affect any of these processes. As

95

Unraveling the signaling events leading to GT-induced JNK activation shown in Figure 9.6, JNK was still similarly phosphorylated at Thr183/Tyr185 and Bim displayed the same gel shift in lysates of BEAS cells treated with GT and Axon-1126 as compared to lysates from cells treated with GT only (Figure 9.6). Nevertheless, cotreatment of GT with Axon-1126 blocked the processing of PARP and caspase-3 cleavage suggesting that the protection from GT-induced cell death by GSK3 inhibition was independent of the JNK/Bim phosphorylation/activation axis.

Figure 9.6 | GSK3 does not act upstream of JNK during GT-induced cell death. Whole cell lysates from BEAS-2B cells treated with 1 µM GT, 0.75 µM GSK3 inhibitor Axon-1126, or both were subjected to immunoblotting. JNK was still phosphorylated at Thr183/Tyr185 in the presence of the inhibitor. Similarly, Bim still showed gel mobility shifts indicating that Bim was also phosphorylated under GSK3 inhibiting conditions. However, employment of Axon-1126 blocked the cleavage of PARP and caspase-3, underlining its anti-apoptotic potential, independent of JNK activation and Bim phosphorylation.

96

Unraveling the signaling events leading to GT-induced JNK activation To examine if GSK3 inhibition conferred resistance to GT by stabilizing the survival factor Mcl-1, we treated Mcl-1 knock-out MEF and Mcl-1 reconstituted variants of these cells with 1 µM GT for up to 6 h and performed Mcl-1 immunoblot analysis. As expected Mcl-1 ko cells were still sensitive to GT-induced caspase-3 processing and PARP cleavage. This was not the case in the Mcl-1 reconstituted cells probably due to the protective effect of overexpressed Mcl-1. Interestingly, the protective effect of the GSK inhibitor Axon-1126 (0.75 µM) was not as strong in Mcl-1 ko cells than previously seen in wt cells. Caspase-3 processing and PARP cleavage were reduced but still clearly detectable upon GT/Axon-1126 treatment of Mcl-1 ko MEF (Figure 9.7, A, compare to Figure 9.6, wt cells). This suggests that GSK3 required the presence of Mcl-1 for its optimal cytoprotective function. Consistent with the notion that this cytoprotection was due to Mcl-1 stabilization, the reconstituted Mcl-1 cells showed less Mcl-1 degradation in response to the combined GT/Axon-1126 treatment than to GT alone (Figure 9.7, A) Similarly, the combination of GT and the GSK3 inhibitor (0.75 µM) stabilized Mcl-1 protein levels in BEAS-2B (Figure 9.7, B). In conclusion, the cell protective effect of the GSK3 inhibitor is likely to be through stabilizing Mcl-1 as Axon-1126 had no inhibitory impact on the state of GTinduced JNK and Bim phosphorylation. It is reasonable to speculate that Axon-1126 stabilized Mcl-1 may better sequester the activated, triple phosphorylated form of Bim than for example Bcl-2 or Bcl-xL (see Figure 7.1) so that phosphorylated Bim is less able to directly activate Bax/Bak.

97

Unraveling the signaling events leading to GT-induced JNK activation

Figure 9.7 | Axon inhibits cell death by stabilizing Mcl-1. Mcl-1 knock-out MEF (Mcl-1 ko) or knock-out MEF reconstituted with wild-type Mcl-1 (reconst. MCL wt) were treated with GT (1 µM) for 6 h. (A) Depletion of Mcl-1 reduced the protective effect of Axon-1126 (Ax) as seen by PARP and caspase-3 cleavage in the presence of GSK3 inhibitor (lane 1-3). Mcl-1 expression blocked PARP and caspase-3 processing. Incubation of reconstituted MEF with Axon-1126 decreased the GT-induced degradation of Mcl-1 (lanes 4-6). (B) Co-treatment of GT and Axon-1126 (0.75 µM) increased Mcl-1 protein levels in GT-challenged BEAS-2B cells.

9.3 ROCK is a MAPKK kinase for MKK4/7 9.3.1 Cytoskeletal changes during GT-induced cell detachment In response to GT, BEAS-2B, MEF and primary murine epithelial cells (AEC type II) detached within 1 h. As cell detachment involves cytoskeletal changes, we monitored them during GT treatment by confocal microscopy. BEAS-2B cells were transiently transfected with GFP-tagged paxillin as a marker for focal adhesions and LifeAct (red) to visualize actin microfilaments and actin dynamics (Figure 9.8). As previously reported, actin fibers were anchored to focal adhesions by the adaptor protein paxillin. After GT treatment, the actin cables (stress fibers) contracted within 40 min but stayed attached to focal adhesions, as seen by co-localization with GFP-paxillin (Figure 9.8). Actin de-polymerization was not observed.

98

Unraveling the signaling events leading to GT-induced JNK activation

Figure 9.8 | Actin fibers contract during detachment. GT-induced cell detachment was monitored by confocal live cell microscopy. BEAS-2B were co-transfected with GFP-paxillin (green) as a marker for focal adhesions and LifeAct (red) to visualize changes of the actin cytoskeleton. Paxillin was located at focal adhesions and anchored to actin fibers. Actin stress fibers contracted, after treatment with 1 µM GT but stayed attached to focal adhesions (co-localization with GFP-paxillin). Inlets (right panel) illustrate that actin filaments contracted but did not de-polymerize. Scale bar = 10 µm.

99

Unraveling the signaling events leading to GT-induced JNK activation Contraction of actin stress fibers and subsequent cell motility and migration events are orchestrated by a variety of kinases and phosphatases. In that context, Rho-associated protein kinase (ROCK) is a key regulator of actin dynamics. Interestingly, this protein kinase has also been linked to apoptosis but the molecular mechanisms of its cell death signaling function are mainly unknown (chapter 2.3.1).

9.3.2 Pharmacological inhibition of ROCK protects against GT As shown above, the JNK activating kinases MKK4 and MKK7 were phosphorylated in response to GT. In addition, we observed the contraction of actin stress fibers in GT-treated BEAS-2B cells pointing towards a possible involvement of actin dynamics regulated by ROCK. We therefore tested if ROCK was the kinase that transduced the apoptotic GT-signaling to MKK4, MKK7 and JNK. We first examined the effect of two pharmacological ROCK inhibitors, H-1152 and Y-27632 on GT-induced apoptosis in BEAS-2B cells. In contrast to untreated cells, ROCK inhibitor treated BEAS-2B cells did not show any visible apoptotic morphology after 6 h of 1 µM GT treatment (Figure 9.9). Strikingly the H-1152-treated cells did not even detach if challenged with GT. Similarly, BEAS-2B depleted of ROCK by shRNAmediated knock-down also remained attached and did not show the formation of membrane blebs (Figure 9.9, right panel).

Figure 9.9 | Inhibition of ROCK restored cell morphology. BEAS-2B treated with the ROCK inhibitor H-1152 or depleted of ROCK1 maintained epithelial morphology when treated with GT. These cells did not detach in response to GT and did not show morphological changes associated with apoptosis as compared to BEAS-2B without ROCK inhibitor treatment. Scale bars = 200 µm.

100

Unraveling the signaling events leading to GT-induced JNK activation The protective effect of ROCK inhibition was confirmed by cell death assays. Both ROCK inhibitors, H-1152 and Y-27632 (1 µM each), prevented caspase-3 activation in BEAS-2B cells, treated with GT for 6 h (Figure 9.10, A).In addition, annexin-V staining was effectively reduced by ROCK inhibition, and this was as efficient as with the general caspase inhibitor QVD (Figure 9.10, B) indicating that ROCK was essential for GT-induced caspase-3 activation and apoptosis.

Figure 9.10 | Cell death was rescued by ROCK inhibition. BEAS-2B cells were treated with 1 µM GT for 6 h. (A) Inhibition of ROCK by two ROCK inhibitors, H-1152 and Y-27632 (1 µM each), significantly rescued GT-induced activation of caspases. (B) Employment of the ROCK inhibitors additionally reduced apoptotic cell death, measured by annexin-V-FITC. Importantly, both ROCK inhibitors rescued GT-induced apoptosis to the same extent as caspase inhibition by QVD (20 µM). Graphs show the means of at least three independent experiments ± S.E.M; p-values: *** < 0.001, oneway anova, posthoc: Bonferroni compared to GT. Importantly, ROCK inhibition did not only protect cells from GT-induced apoptosis but also maintained the epithelial morphology. Next, we sought to investigate if ROCK mediated the GT-induced caspase-3 activation and apoptosis by modulating JNK and Bim phosphorylation/activity.

101

Unraveling the signaling events leading to GT-induced JNK activation

9.3.3 ROCK relays GT signaling to MKK4/7 and JNK For that purpose, we treated BEAS-2B cells with or without GT and ROCK inhibitors for 6 h, prepared whole cell lysates and subjected them to immunoblot analysis using phosphospecific antibodies against the active forms of MKK4/7 and JNK. The cell death protective properties of H-1152 and Y-27632 were confirmed by the fact that both ROCK inhibitors ablated GT-induced caspase-3 processing and drastically

diminished

PARP

cleavage

(Figure

9.11, A).

Moreover,

JNK

phosphorylation and hence activation was not detectable after co-treatment of the cells with GT and the ROCK inhibitors. As expected QVD did not affect JNK phosphorylation but blocked GT-induced caspase-3 processing and cell death.

Figure 9.11 | ROCK acts upstream of JNK and Bim. Immunoblots of BEAS-2B cells treated with GT (1 µM) for 6 h and the ROCK inhibitors H-1152 (1 µM) or Y-27632 (1 µM). (A) GT-induced the activating phosphorylation of JNK after 6 h. Combination with ROCK inhibitors blocked phosphorylation of JNK as well as the induction of cell death, as monitored by reduced PARP and caspase-3 cleavage. Caspase inhibition with QVD (20 µM) protected from cell death induction but did not affect JNK phosphorylation.

(B)

Diminished

JNK

activation

correlated

with

reduced

phosphorylation of Bim. Bim phosphorylation was assessed by pT112/pS114 specific antibody immunoblotting (upper panel) or by gel mobility shift analysis (lower panel). 102

Unraveling the signaling events leading to GT-induced JNK activation With regard to Bim, the combined treatment of GT (1 to 2 µM) with the ROCK inhibitor H-1152 (1 µM) reduced both the gel mobility shift and the phosphorylation of Bim at S112/T114 which is usually detected by GT treatment alone (Figure 9.11, B). Reduced phosphorylation of JNK after ROCK inhibitor treatment correlated with reduced JNK activity as assessed by immunoprecipitating JNK from whole cell lysates and subsequent in vitro phosphorylation of recombinant c-Jun on the pulldowns. Quantification of anti-phospho c-Jun immunoblots from these pulldowns (Figure 9.12, A) revealed that H-1152 diminished GT-induced c-Jun phosphorylation to the same level as in untreated cells (Figure 9.12, B). This shows that ROCK was crucial for GT-induced JNK activation and c-Jun phosphorylation.

Figure 9.12 | ROCK mediated GT-induced c-Jun phosphorylation. BEAS-2B cells were treated with 1 µM Gliotoxin (GT) for 6 h in absence or presence of the ROCK inhibitor H-1152. (A) JNK was enriched from whole cell lysates and incubated with recombinant cJun. Phosphorylated cJun was detected by immunoblotting. (B) Quantification showed reduced JNK activity in cells co-treated with ROCK inhibitor and GT. Graphs show the means of at least three independent experiments ± S.E.M; pvalues: *0.05-0.01, one-way anova, posthoc: Bonferroni compared to GT.

103

Unraveling the signaling events leading to GT-induced JNK activation ROCK was also required for JNK and Bim phosphorylation/activation and caspase-3 activation in MEF since the addition of the ROCK inhibitor (1 µM) reduced GT-induced JNK phosphorylation, Bim gel mobility shifts and phosphorylation at T112/S114 as well as caspse-3 processing (Figure 9.13). In addition, ROCK inhibition also reduced the phosphorylation of MKK4 indicating that ROCK activity was upstream of this protein kinase.

Figure 9.13 | ROCK is required for GT-induced MKK4, JNK and Bim phosphorylation leading to caspase-3 activation. Reduced phosphorylation of MKK4, JNK and Bim as well as decreased caspase-3 processing were seen in MEFs co-treated with 1 µM GT and the ROCK inhibitor H-1152 (1 µM). Phosphorylation of Bim was analyzed by pT112/pS114 specific antibody (pBim) and by gel mobility shifts (Bim).

104

Unraveling the signaling events leading to GT-induced JNK activation

9.3.4 Downregulation of ROCK confirms pharmacological cytoprotection As shown above, pharmacological inhibition of ROCK protected cells from GT-induced JNK activation and subsequent apoptosis. We next examined if a shRNA-mediated depletion of ROCK1 would give the same result. Figure 9.14, A shows that we could achieve a complete knock-down of ROCK1 expression by lentiviral shRNA transfer into BEAS-2B cells. As with the ROCK inhibitors, the cells lacking ROCK1 expression showed reduced phosphorylation of JNK and caspase-3 processing in response to GT (Figure 9.14, B). BEAS-2B cells with a shRNA construct targeting luciferase were used as controls and showed the expected GT-induced JNK phosphorylation and caspase-3 cleavage.

Figure 9.14 | ROCK1 knock-down confirms the finding with pharmacological ROCK inhibitors. (A) Immunoblot showing the successful depletion of ROCK1 by lentiviral shRNA transfer in BEAS-2B cells. (B) These cells showed reduced phosphorylation of JNK and subsequent caspase-3 processing as compared to control cells infected with a shRNA against luciferase (shCTRL). Cells were treated with GT for 6 h. Consistent with the reduced caspase-3 processing, ROCK1 knock-down BEAS2B cells exhibited a diminished caspase-3 activity in their lysates after GT treatment

105

Unraveling the signaling events leading to GT-induced JNK activation (Figure 9.15, A). This was associated with lower annexin-V staining (Figure 9.15, B) indicating that cells lacking ROCK1 did not undergo GT-induced apoptosis as much as wt cells.

Figure 9.15 | Depletion of ROCK interferes with GT-induced apoptosis. BEAS-2B depleted for ROCK (shROCK1) and control cells infected with a shRNA against luciferase (shCTRL), were treated with 1 µM GT for 6 h. (A) Cells lacking ROCK1 expression showed significantly reduced activation of caspases. (B) Reduced caspase activity correlated with a significantly decreased number of annexin-V positive cells. Graphs show the means of at least three independent experiments ± S.E.M; p-values: *** < 0.001, one-way anova, posthoc: Bonferroni compared to shCTRL. In summary, we showed by two different methods, pharmacological inhibition and shRNA-mediated knock-down, that ROCK is a key player in the transmission of GT signaling to the cytoskeleton and to activate the apoptotic JNK-Bim pathway. Inhibition or lack of expression of ROCK prevents GT-induced MKK4, JNK and Bim phosphorylation and blunts caspase-3 processing/activation and consequently apoptosis of lung epithelial cells. In addition, it preserves the attached epithelial morphology of the cells.

106

Unraveling the signaling events leading to GT-induced JNK activation

9.4 RhoA is activated upstream of ROCK and mediates apoptosis For its activation, ROCK is relieved from an auto-inhibitory conformation by the upstream activator Rho, which binds to a particular Rho-binding domain of ROCK. The proteins of the Rho family (Rho, CDC42 and Rac) are small GTPases that are activated upon GDP-GTP exchange. Here, two bacterial toxins were employed to study the involvement of Rho proteins in transmitting the apoptotic signaling of GT to ROCK. The two-component fusion toxin C3FT (C2II toxin from Clostridium botulinum and C3 exoenzyme from Clostridium limosum)234 inhibits Rho proteins by stabilizing the interaction of Rho-GDP with GDP dissociation inhibitors (GDI), therefore restraining Rho activation. In contrast, the bacterial toxin CNFy from Yersinia pseudotuberculosis, inhibits the GTPase activity of Rho proteins and thereby prolongs their active GTP-bound state (Figure 9.16, A).

Figure 9.16 | GT activates RhoA. (A) Two toxins were employed to monitor Rho activity. The C3 fusion toxin prevents Rho activation by stabilizing the binding to GDIs. CNFy toxin blocks the GTPase activity of Rhos and thus keeps Rho proteins active. (B) Active Rho proteins were enriched by Rhotekin pulldown of lysates from BEAS-2B cells, treated with GT for the indicated time points. Immunoblots were probed with RhoA specific antibodies. Active RhoA was detected starting from 40 min post GT treatment. The Rho activating toxin CNFy (150 ng/ml) was applied for 4 h to BEAS-2B cells as a positive control. RhoA activity was assessed by Rhotekin pull-down. Rhotekin is a Rho binding domain containing effector protein of all Rho proteins and hence interacts with their active GTP-bound state. Rhotekin-coupled to agarose beads therefore allowed for the

107

Unraveling the signaling events leading to GT-induced JNK activation enrichment of active Rho’s from whole cell lysates. We detected a gradual increase in active RhoA-GTP after 20-40 min of GT-treatment, when we applied these beads to lysates of GT-treated BEAS-2B cells and tested the pulldowns with anti-RhoA antibodies on immunoblots (Figure 9.16, B). Cells treated for 4 h with CNFy (150 ng/ml) served as a positive control for RhoA activation. The increase in RhoA activity after GT treatment correlated with the activation of ROCK. For that purpose we immunoblotted the lysates of BEAS-2B cells treated with 1.5 µM GT for up to 4 h with a phosphospecific antibody against a major substrate of ROCK, the myosin-binding subunit of myosin phosphatase 1 (MYPT1). MYPT1 phosphorylation was detected after 1 h of GT treatment and increased further for up to 4 h (Figure 9.17, A). Pre-treatment of cells with the Rho inhibitory toxin C3 (200 ng/ml C2II and 100 ng/ml C3FT) for 1 h drastically reduced the levels of phosphorylated MYPT1, confirming that GT-triggered ROCK activation via Rho. Figure 9.17, B shows a quantification of the anti-pMYPT1 Western blot illustrating this point.

Figure 9.17 | Inhibition of RhoA blocks GT-induced ROCK activity. BEAS-2B cells were treated with 1.5 µM GT for the indicated time points. Cells were either directly challenged with GT or preincubated with C3 toxin prior to the combined treatment of C3 and GT. (A) Immunoblots of whole cell lysates were probed with an antibody against phosphorylated MYPT1 (pT696). MYPT1 was phosphorylated within 1 h of GT exposure, and this phosphorylation increased further during the treatment. Inhibition of Rho (C3) reduced MYPT1 phosphorylation. (B) Blots were quantified to assess the ROCK activity by calculating the ratio of pMYPT1 to total MYPT1. Graphs show the means of at least three independent experiments ± S.E.M; p-values: * 0.05-0.01, oneway anova, posthoc: Bonferroni compared to GT. 108

Unraveling the signaling events leading to GT-induced JNK activation Inhibition of Rho by C3 toxin not only decreased ROCK activity but also reduced the detachment of BEAS-2B cells in response to GT (Figure 9.18), similar to what we have seen with the direct inhibition of ROCK (Figure 9.9). Cells were pre-treated for 1 h with Rho inhibiting (C3) or Rho activating toxins (CNFy), before being stressed with GT (1 µM). Only the combined treatment with GT and C3 toxin, but not CNFy, reduced the detachment of BEAS-2B (Figure 9.18).

Figure 9.18 | C3 toxin rescues epithelial morphology. BEAS-2B cells were either pre-treated in full media, media containing C3FT, or CNFy for 1 h before adding GT for 2 h. Inhibition of Rho by C3 and hence reduced ROCK activity prevented cell detachment. Scale bar = 100 µm. In line with the reduced ROCK activity, the application of C3 toxin diminished the GT-induced phosphorylation of MKK4, JNK and Bim. After preincubating BEAS-2B cells with the C3 toxin for 1 h, the GT-induced phosphorylation of MKK4 at S257/T261 was significantly reduced (Figure 9.19). This led to diminished JNK phosphorylation at T183/Y185 as well as Bim phosphorylation at T112/S114 (Figure 9.19, B).

109

Unraveling the signaling events leading to GT-induced JNK activation

Figure 9.19 | Rho activity is upstream of MKK4 and JNK. BEAS-2B cells were treated with GT and the Rho inhibiting (C3) or RhoA activating (CNFy) toxins. (A) Whole cell lysates were analyzed using phospho-specific antibodies. Co-treatment of GT with C3 toxin reduced phosphorylation of MKK4 (pS257/pT261) as well as JNK phosphorylation (pT183/pY185) after 6 h. (B) Co-treatment with C3 toxin and GT also blocked phosphorylation of Bim (pT112/pS114). Blots were brightness adjusted for visibility. In summary, we found that RhoA was activated within 40 min after GT treatment. Inhibition of Rho proteins by the bacterial toxin C3 rescued GT-induced cell detachment, decreased ROCK activity, and consequently reduced phosphorylation, and thus activation, of the MKK-JNK-Bim signaling axis.

110

Unraveling the signaling events leading to GT-induced JNK activation

9.4.1 Inhibition of CDC42 and Rac1 does not protect from apoptosis The Rho family is characterized by a highly interconnected signaling network between the different family members. A possible involvement of the Rho family members CDC42 and Rac1 was investigated by using pharmacological inhibitors. ML-141 inhibits both CDC42 and Rac1 whereas NSC-23766 specifically inhibits Rac1 GEFs. In contrast to the C3 toxin, the co-treatment of BEAS-2B cells with 1 µM GT and increasing concentrations of ML-141 (1-100 µM) did not inhibit cell detachment, and the cells still exhibited the membrane blebs characteristic of apoptosis (Figure 9.20).

Figure 9.20 | Inhibition of CDC42/Rac1 maintains apoptotic morphology in response to GT. Addition of the CDC42 and Rac1 inhibitor ML-141 to 1 µM GT did not prevent the formation of apoptotic membrane blebs. Scale bar = 100 µm. The influence of Rac1 and/or CDC42 on the GT-mediated induction of the proapoptotic JNK-Bim pathway was also analyzed by immunoblotting. Increasing concentrations of ML-141 did not reduce the T183/Y185 phosphorylation of JNK or the T112/S114 phosphorylation of Bim (Figure 9.21). JNK/Bim phosphorylation/activation was

also

unchanged

using

the

Rac1

specific

inhibitor NSC-23766.

The

phosphorylation specific signals for JNK and Bim were rather increased, if the latter inhibitor was combined with GT.

111

Unraveling the signaling events leading to GT-induced JNK activation

Figure 9.21 | CDC42/Rac1 inhibition do not prevent activation of the JNK-Bim pathway. Co-treatment of CDC42/Rac1 inhibitors with 1 µM GT for 6 h showed similar phosphorylations and hence activation of JNK (T183/Y185) and Bim (T112/S114). Bim phosphorylation was also seen by gel mobility shifts. Both, gel shifts and the pT112/pS114 Bim specific signals were even increased using NSC-23766. Thus, since the inhibition of Rac1 and CDC42 did not block the induction of the GT-induced JNK-Bim signaling pathway we conclude, that these two classes of GTPases are not involved in GT-induced apoptosis. We suggest that rather RhoA is the major component in transducing the apoptotic signals from GT to ROCK activation and Bim-mediated Bax/Bak activation leading to subsequent cell death.

112

Unraveling the signaling events leading to GT-induced JNK activation

9.5 GT triggers cell death by interfering with cell adhesion signaling Adherent cells treated with GT detached before morphological signs of cell death could be observed (Figure 9.22). This suggested that integrins, the major surface proteins that mediate cell adhesion to the extracellular matrix and the associated focal adhesion complex, might be involved in GT-induced apoptosis.

Figure 9.22 | GT detaches cells before the induction of apoptosis. Primary, murine alveolar epithelial cells (AEC type II) detached within 1 h after GT treatment. No morphological indications for apoptosis could be observed at this time point. Scale bar = 100 µm. The focal adhesion complex anchors the cytoskeleton to integrins. At the cytosolic tail of integrins it consists of a variety of adaptor proteins and kinases (see chapter 2.3.2). The purpose of this complex is to relay pro-survival signaling when integrins engage the extracellular matrix and to coordinate cell adhesion and migration. Here we sought to study if GT targeted this complex in order to provoke cell detachment and subsequent cell death.

113

Unraveling the signaling events leading to GT-induced JNK activation

9.5.1 GT translocates paxillin from focal adhesions to early endosomes GFP-labeled paxillin was used to visualize the dynamics of focal adhesions in GT-treated cells. Paxillin is a scaffold protein at focal adhesions and involved in the recruitment of focal adhesion kinase (FAK), the essential signaling enzyme in the complex. BEAS-2B lung epithelial cells overexpressing GFP-paxillin were monitored by confocal live cell microscopy after GT challenge. Paxillin translocated from a focal adhesion pattern to a vesiclular phenotype within the first minutes of GT action (Figure 9.23). The inward movement of paxillin containing vesicles clearly occured prior to cell detachment, typically within 45 min post GT addition or faster, indicating that this process is mechanistically linked to detachment rather than a downstream side effect.

Figure 9.23 | Paxillin translocates to a vesicular pattern before cell detachment. GT-induced cell detachment was followed by confocal live cell imaging using BEAS-2B

114

Unraveling the signaling events leading to GT-induced JNK activation cells transfected with GFP-paxillin. Paxillin translocated from the typical focal adhesion staining (top left) to a vesicular pattern shortly after GT treatment. Scale bar = 10 µm. To identify the nature of the vesicular paxillin localization, BEAS-2B cells were co-transfected with GFP-paxillin and an endosomal marker (mRuby-Endo-14). Again, paxillin rapidly translocated from a focal adhesion pattern in response to GT treatment and localized, at least in part, to endosomal membranes before the cell detached (Figure 9.24).

Figure 9.24 | Paxillin co-localizes with early endosomes. Co-transfection of BEAS2B cells with GFP-paxillin (green) and mRuby-Endo-14 (red) to visualize endosomes. Paxillin partially localized to endosomal membranes. Inlets mark the region of digital magnification

to

illustrate

the

localization

of

paxillin

(bottom

pictures).

Scale bars = 10 µm.

115

Unraveling the signaling events leading to GT-induced JNK activation

Next, we examined paxillin protein levels in the lysates of GT-treated BEAS-2B cells in order to elucidate the nature of this paxillin translocation to endosomes. Anti-paxillin immunoblotting of these lysates showed that the levels of paxillin rapidly decreased within 30 min of GT treatment, and they were not detectable anymore after 2 h (Figure 9.25).

Figure 9.25 | GT induces the degradation of paxillin. BEAS-2B cells were treated with 1 µM GT for the indicated time points and the cell lysates were subjected to antipaxillin immunoblotting. Paxillin protein levels decreased within 30 min after addition of GT and became undetectable after 2 h. Taken together, these data suggest that focal adhesions considerably change their structure and protein composition in response to GT. Endocytosis and loss of the paxillin protein imply alterations in FAK signaling. We therefore sought to identify the role of FAK in GT-induced apoptosis.

9.5.2 FAK and p190RhoGAP are both inactivated by GT FAK is an interesting target of GT in the context of Rho signaling because its activation is known to suppress the Rho family during stress fiber formation and focal adhesion turn-over177,235. These findings raised two important questions (i) can FAK regulate RhoA-dependent cell death and (ii) how is FAK regulated in response to GT. When FAK is recruited to growing focal adhesions due to the attachment of cells to the extracellular matrix, it gets activated by autophosphorylation at Y397 (see chapter 2.3.2). This activating phosphorylation can be monitored by phosphospecific antibodies. As shown in Figure 9.26, FAK was indeed phosphorylated at Y397 in untreated, attached BEAS-2B cells. However, in response to GT, this activating

116

Unraveling the signaling events leading to GT-induced JNK activation phosphorylation was rapidly lost within 30 min, which is concomitant with the time of cell detachment (Figure 9.26).

Figure 9.26 | FAK is rapidly dephosphorylated in response to GT. Immunoblots of lysates of BEAS-2B cells treated with 1 µM GT for the indicated time points were either probed with an antibody against the activating phosphorylation of FAK at Y397 (pFAK) or an antibody against total FAK. FAK phosphorylation decreased within 30 min after administration of GT. The decreased phosphorylation of FAK at Y397 indicated that FAK was inactivated in response to GT. Active FAK is known to phosphorylate the focal adhesion components paxillin and p190RhoGAP (p190). We therefore monitored the phosphorylation status of these two proteins in the absence or presence of GT treatment by immunoblotting them with phosphospecific antibodies. In response to GT, paxillin was dephosphorylated at Y118 with the same time kinetic as FAK dephosphorylation/inactivation (Figure 9.27, A). In addition, paxillin was rapidly degraded, probably due to its endocytosis and subsequent lysosomal degradation.

Importantly,

paxillin

dephosphorylation

and

degradation

were

concomitant with the phosphorylation and activation of JNK (Figure 9.27, A). p190 is a GTPase activing protein for RhoA in focal adhesions and it is activated by FAK-dependent phosphorylation. Phosphorylated, active p190 therefore favors the formation of the GDP-bound, inactive state of RhoA. Conversely dephosphorylated, inactive p190 allows a better activation of RhoA. As shown in Figure 9.27, B, GT treatment of BEAS-2B cells led to the dephosphorylation and inactivation of p190 within 1 h, which then should facilitate RhoA activation.

117

Unraveling the signaling events leading to GT-induced JNK activation

Figure 9.27 | Paxillin and p190RhoGAP are rapidly dephosphorylated and inactivated together with FAK by GT. BEAS-2B cells were subjected to immunoblot analysis to monitor the phosphorylation status of downstream targets of FAK after treatment with 1 µM GT for the indicated time points. (A) Phosphorylation of paxillin at Y118 (pPax) was reduced as paxillin protein levels decreased. Importantly JNK phosphorylation at T183/Y185 (pJNK) occurred at the same time. (B) The FAK-dependent Y1105 phosphorylation of p190RhoGAP (pP190) also decreased in response to GT with a similar kinetic as FAK dephosphorylation/inactivation. Importantly, the kinetic of dephosphorylation of FAK, paxillin and p190 were perfectly in line with the observed activation of RhoA after 40 min of GT treatment (Figure 9.16). Our data therefore suggest that GT inactivated FAK and subsequently prevented the activating phosphorylation of p190RhoGAP, thereby stimulating Rho activity. To further investigate if inactivation of FAK represents the trigger to activate the pro-apoptotic pathway from RhoA-ROCK to JNK, we employed two different pharmacological FAK inhibitors. Indeed, the FAK inhibitor PF-562-271 (1 µM) induced caspase activity as a single agent in BEAS-2B cells (Figure 9.28, A). Moreover, Rhotekin pulldown assays showed that the inhibitor could activate RhoA in the absence of GT (Figure 9.28, B). However, the kinetic of these events was considerably different from that of GT-induced cell death. While GT activated RhoA within 40 min and caspase-3 within 4-6 h (see Figure 9.4 and Figure 9.16), the FAK inhibitor started to reveal these changes only after 16 h (Figure 9.28).

118

Unraveling the signaling events leading to GT-induced JNK activation

Figure 9.28 | FAK inhibition induces RhoA activity and cell death. BEAS-2B cells were challenged with the FAK inhibitor PF-562-271 (1 µM) for the indicated time points. (A) Caspase activity was induced in whole cell lysates starting from 16 h after addition of the inhibitor. (B) Rhotekin pull-downs were performed to enrich active, GTP-bound Rho proteins from whole cell lysates. Immunoblotting, using a RhoA specific antibody, showed activation of RhoA by direct inhibition of FAK without additional GT treatment. Graphs show the means of at least three independent experiments ± S.E.M; p-values: ** 0.01-0.001, *** < 0.001, one-way anova, posthoc: Bonferroni compared to nontreated (NT, 72h). Along the signaling pathway, the FAK inhibitor PF-562-271 diminished both the phosphorylation of FAK at Y397 as well as the activating Y1105 phosphorylation of p190, similar to what we have seen with GT but again at later time points (16-24 h, Figure 9.29, A). Since p190 phosphorylation decreased, ROCK activity was increased as assessed by increased phosphorylation of its substrate MYPT1. In addition, activation of ROCK well correlated with JNK activation/phosphorylation at T183/Y185 as well as with caspase-3 processing (Figure 9.29, A). The second FAK inhibitor (FAK 14, 50 µM) confirmed our data, now even with time kinetics similar to those of GT. As shown in Figure 9.29, B, the inhibitor also inactivated FAK and p190RhoGAP and stimulated MKK4 (pS257/pT261) and JNK (pT183/pY185) phosphorylations as efficiently as GT. Thus, FAK inhibition seems to mimic the GT-induced apoptosis signaling pathway via p190RhoGAP inactivation and subsequent RhoA/ROCK and MKK/JNK phosphorylation/activation.

119

Unraveling the signaling events leading to GT-induced JNK activation

Figure 9.29 | Inhibition of FAK mimics GT signaling. Two different FAK inhibitors were employed to analyze activation of the RhoA/ROCK dependent apoptotic pathway by immunoblotting. (A) The FAK inhibitor PF-562-271 (1 µM) decreased the activating phosphorylations of FAK (pY397, pFAK) and p190RhoGAP (pY1105, pp190) in BEAS-2B cells. ROCK associated MYPT1 phosphorylation (pT696, pMYPT1) increased after inactivation of the focal adhesion complex. JNK phosphorylation (pT183/pY185, pJNK) was observed simultaneously and its activation correlated with caspase-3 processing. (B) The inhibitor FAK14 (50 µM) reduced pFAK and pp190 intensities with a kinetic similar to GT. Inactivation of the focal adhesion proteins after 30 min was in line with the detection of activating phosphorylations for MKK4 and JNK.

120

Unraveling the signaling events leading to GT-induced JNK activation To investigate if high levels of FAK could modify the apoptotic response to GT we overexpressed wild-type FAK in BEAS-2B by lentiviral transduction (Figure 9.30, A). Surprisingly, GT was even able to dephosphorylate and inactivate overexpressed FAK. Nevertheless, the remaining abundance of overexpressed, active FAK was able to reduce the GT-induced phosphorylation of JNK leading to a minor diminishment of caspase-3 activity as compared to empty vector (e.v.) control cells (Figure 9.30, B).

Figure 9.30 | FAK overexpression partially protects BEAS-2B cells from GT-induced JNK phosphorylation and caspase-3 activation. (A) BEAS-2B cells were lentivirally infected to express FAK at high levels (FAK oe). Cells infected with empty vector (e.v.) were used as a control. GT (1 µM) efficiently inactivated FAK in both conditions. Nevertheless, JNK phosphorylation (pT183/pY185) was still decreased in FAK oe cells. (B) High levels of FAK provided partial protection from the activation of caspases after 6 h of GT treatment as compared to control cells. Graphs show the means of at least three independent experiments ± S.E.M; p-values: ** 0.010.001, two-way anova, posthoc: Bonferroni compared to e.v. (GT-treated). Taken together, our data show that (i) FAK is rapidly inactivated by GT and (ii) by this inactivation it can induce the RhoA-dependent, pro-apoptotic kinase cascade in a similar manner as GT. 121

Unraveling the signaling events leading to GT-induced JNK activation

9.5.3 GT interferes with the integrin binding capacity to the ECM GT was shown to rapidly induce cell detachment and inactivation of proteins of the focal adhesion complex. Therefore, it is reasonable to speculate that integrins might be targeted or affected by this toxin. GT has been postulated to exert its toxicity through covalent modification of cysteine thiol groups210. Integrins are interesting GT targets in this regard, because integrin activity and ligand binding is, at least in part, regulated by a divers network of disulfide bridges between intramolecular cysteines (chapter 2.3.2). Recombinant human integrin αVβ3 (5 µg) was incubated with GT (1 mM) overnight and subjected to mass spectrometry in order to detect possible modifications of cysteines. Indeed, we detected the covalent interaction of GT with particular cysteines in both analyzed integrin chains. Integrin αV was modified at Cys158 in the second repeat of the seven blade β-propeller domain that mediates ligand binding (Figure 9.31, A). Similarly, integrin β3 was modified in the ligand binding head domain at Cys258 (Figure 9.31, B). Interestingly, both αV Cys158 and β3 Cys258 are highly conserved among the respective subunits (Figure 9.32).

122

Unraveling the signaling events leading to GT-induced JNK activation

Figure 9.31 | GT covalently modifies recombinant integrin αVβ3. Recombinant human integrin αVβ3 (5 µg) was digested with trypsin. Peptides were incubated overnight with GT (1 mM). Peptides were analyzed by MS and identified by Maria Koczorowska (AG Schilling) using the Trans-Proteomic Pipeline (TPP) and mapped to the protein sequence using the software Proteator (unpublished, AG Schilling)233. Protein sequences were retrieved from Uniprot. Identified peptides were mapped in grey. Peptides with the additional mass of GT were mapped in blue. (A) The αV chain showed binding of GT at Cys158 in the seven blade β-propeller domain (brown). (B) The β3 chain was modified at Cys258 in the ligand binding domain (b-I). Domain structures were drawn using the Illustrator for Biological Sequences (IBS V1.0, not in scale). PSI: plexin-semaphorin-integrin domain, Hyb: Hybrid domain, I-IV: I-EGF domains 1 to 4, β-T: beta tail.

123

Unraveling the signaling events leading to GT-induced JNK activation

Figure 9.32 | GT-targeted cysteines are conserved. Protein sequences were retrieved from Uniprot as indicated on the left (http://www.uniprot.org/) and aligned using Clustal Omega (V1.2.1)236,237. (A) Cys158 (relative to human αV, green box) was conserved in all α subunits. Interestingly, this cysteine forms a disulfide bridge to Cys138. (B) Similarly, the Cys258 of the β3 subunit (green box) was also conserved and is linked to Cys299 (not shown) by a disulfide bond. Disulfide linkage was first described elsewhere166. We next reasoned that if the binding of GT to active integrins in their extracellular binding site is the way how it induces apoptosis of adherent cells, it should not kill suspension cells because their integrins are in an inactive state with their binding site hidden (Figure 2.12, A). We therefore compared the adherent lung epithelial cells BEAS-2B with a panel of suspension cells to test the activation states of their integrins and their sensitivity to GT-induced apoptosis. Apoptosis induction was assessed by annexin-V staining after 6 h treatment with GT. All tested suspension cell lines (BAF3, Jurkat, FL5.12) were less sensitive to GT-induced apoptosis as compared to adherent cells (Figure 9.33, A). BAF3 cells were subjected to immunoblotting to investigate activation of the pro-apoptotic kinase

124

Unraveling the signaling events leading to GT-induced JNK activation cascade. Consistent with their resistance to GT-induced apoptosis, neither MKK4 (pS257/pT261) nor JNK (pT183/pY185) phosphorylations were detectable in these suspension cells after GT treatment (Figure 9.33, B).

Figure 9.33 | Suspensions cells are not sensitive to GT. Adhesion (BEAS-2B) and suspension cells were treated for 6 h with 1 µM GT to monitor a possible influence of integrins on the transmission of GT signaling. (A) The induction of apoptosis was assessed by annexin-V staining. Murine suspension cells (BAF3 and FL5.12), as well as human suspension cells (Jurkat), were less sensitive to GT-induced cell death. (B) Protection from cell death induction correlated with absence of detectable MKK4 and JNK phosphorylation. In contrast, phosphorylation of MKK4 at S257/T261 and JNK phosphorylation at T183/Y185 were present in adherent BEAS-2B cells (left lane). Graphs show the means of at least three independent experiments ± S.E.M; p-values: *0.05-0.01, two-way anova, posthoc: Bonferroni compared to NT. We next assessed the fate of integrins after GT stimulation in adherent and suspension cells by FACS analysis using specific anti-human β3 and anti-mouse β1 integrin antibodies. All cells showed a marked expression of integrin β chains on their surface. However, while adherent BEAS-2B cells lost integrin β3 surface expression within 4-6 h after GT treatment (1 µM) this was not the case with the Jurkat suspension cell line (Figure 9.34, A). Similarly, mouse embryonic fibroblasts diminished their surface integrin β1 expression following GT treatment whereas the BAF3 and FL5.12

125

Unraveling the signaling events leading to GT-induced JNK activation suspension cells maintained their integrins on the surface (Figure 9.34, B). Representative histograms of the FACS staining are shown in Appendix 2.

Figure 9.34 | GT decreases the surface expression of integrins on adherent, but not suspension cells. Surface expression of human integrin β3 or murine integrin β1 was assessed by FACS staining, using specific antibodies. All cells were treated for the indicated time points with 1 µM GT. (A) Levels of integrin β3 were reduced on the surface of BEAS-2B cells after 2 h. In contrast, Jurkat cells did not show a significantly reduced surface expression. (B) Similarly, murine adherent cells (MEFs) displayed reduced surface staining for integrin β1, whereas the suspension cell lines BAF3 and FL5.12 maintained the initial levels. Graphs show the means of at least three independent experiments ± S.E.M; p-values: *0.05-0.01, ** 0.01-0.001, *** < 0.001, two-way anova, posthoc: Bonferroni compared to NT. The data so far indicated that GT had only a significant effect on adherent cells. It stimulated the removal of integrins from the cell surface and induced apoptosis. None of these events were seen with suspension cells although they expressed integrins on their surface as well (but most likely in an inactive, inaccessible state). Our next step was to identify how exactly GT affected integrin function. The removal of integrins from the surface was a rather slow process (4-6 h) as compared to the impact of GT on FAK inactivation and RhoA activation (40 min, see Figure 9.16 and Figure 9.26). It was therefore likely that GT stimulated the endocytosis of integrins but that this process is rather a consequence of a more rapid integrin inactivation by GT.

126

Unraveling the signaling events leading to GT-induced JNK activation For that reason, integrin activity was studied in more detail. Active integrins are known to bind to the RGD motif of fibronectin. We therefore used RGD peptides labeled with the fluorophore 5-FAM to visualize the activation status of surface integrins by FACS analysis. As shown in Figure 9.35, integrins were effectively labeled with RGD-5-FAM on the surface of untreated BEAS-2B cells indicating that they were in an active state to bind fibronectin. In response to GT, the labeling however disappeared within 30 min to 2 h and after 4 h it partially resumed. Importantly, the kinetic of the loss of RGD binding by GT perfectly coincided with that of FAK inactivation and RhoA activation. Only little RGD-5-FAM binding was noted for suspension cells (Figure 9.37). This finding shows that GT interferes with the binding of integrins to fibronectin on adherent cells, probably by covalently modifying particular cysteines in the integrin binding pocket. Suspension cells do not seem to have an exposed fibronectin-binding domain of integrin that can be targeted by GT.

Figure 9.35 | Integrins are inactivated in response to GT. Integrin activity was measured by staining with the fibronectin derived, fluorescently labeled RGD-5-FAM peptide. Fluorescence intensities were followed during GT treatment (1 µM) for the indicated time points by FACS analysis. The fibronectin binding activity of integrins started to decrease after 30 min in response to GT and partially resumed after 4 h. Graphs show the means of at least three independent experiments ± S.E.M; p-values: *0.05-0.01, ** 0.01-0.001, *** < 0.001, two-way anova, posthoc: Bonferroni compared to NT.

127

Unraveling the signaling events leading to GT-induced JNK activation If GT specifically targets active integrins on adherent cells, the artificial attachment of suspension cells to fibronectin should re-activate integrins and make the cells sensitive to GT-induced RhoA/ROCK activation and apoptosis. We therefore plated BAF3 suspension cells on fibronectin-coated plates in order to induce attachment and hence activation of integrins. Indeed, cells attached after an overnight incubation on fibronectin plates (Figure 9.36). Interestingly, these cells were now sensitized to GT, because they rapidly detached and showed the typical apoptotic morphology after GT challenge.

Figure 9.36 | Fibronectin-attached BAF3 cells are sensitive to GT. BAF3 suspension cells were grown overnight on fibronectin-coated plates (50 µg/ml). Attached BAF3 cells were then challenged with 1 µM GT for indicated time points. These cells detached in response to GT and showed apoptotic membrane blebs. Scale bar = 100 µm. To test if the induced attachment of suspension cells activated the fibronectin binding activity of integrins, we performed on these cells the RGD-5-FAM FACS analysis as described above. As show in Figure 9.37, A, the low RGD binding of BAF3 suspension cells increased when they were plated on fibronectin coated plates. Moreover, these cells detached if challenged with GT (Figure 9.36) and underwent 128

Unraveling the signaling events leading to GT-induced JNK activation apoptosis as shown by their membrane blebbing morphology (Figure 9.36) and the positive annexin-V staining (Figure 9.37, B).

Figure 9.37 | Attachment of suspension cells activates their surface integrins for RGD binding and sensitizes them to GT-induced apoptosis. BAF3 cells were attached on fibronectin-coated plates overnight and subsequently challenged with 1 µM GT for 6 h. (A) The attached cells were subjected to RGD-5-FAM peptide staining. Attached cells showed an increased binding to the RGD peptide, indicating integrin activity. (B) Only cells that were positive for RGD binding were able to induce apoptosis as measured by annexin-V staining, 6 h after GT treatment. Graphs show the means of at least three independent experiments ± S.E.M; p-values: *0.05-0.01, Student’s t-test. To determine if the artificially attached, sensitized BAF3 cells now also activated the same GT-induced apoptotic signaling pathway as classical BEAS-2B cells, we again performed a series of immunoblots using phosphospecific antibodies against the protein kinases involved. As previously shown in Figure 9.33, B, suspension BAF3 cells failed to activate the GT-induced pathway that activated Bim via a RhoA/ROCK/MKK4/JNK signaling cascade. However, when BAF3 cells were grown on fibronectin coated plates they exhibited phosphorylation of MKK4 at S257/T261 and JNK phosphorylation at T183/Y185 after treatment with GT for 6 h (Figure 9.38). In addition, we observed an increase in caspase-3 processing during the GT treatment, confirming the induction of apoptosis as shown before by annexin-V staining. 129

Unraveling the signaling events leading to GT-induced JNK activation

Figure 9.38 | Attachment of BAF3 cells to fibronectin coated plates restores their capacity to fully activate MKK4/JNK and caspse-3 in response to GT. Fibronectinattached BAF3 cells were subjected to immunoblotting after treatment with 1 µM GT for the indicated time points. MKK4 was phosphorylated at S257/T261 in attached cells but not in cells that were cultured in suspension. Similarly, T183/Y185 phosphorylation of JNK was increased and this also correlated with caspase-3 cleavage in these attached cells. These experiments show that activation of integrins on suspension cells renders them sensitive to GT and restores the capacity of GT to activate the RhoA-dependent apoptosis pathway. These data therefore support the hypothesis that GT induces apoptosis via the inactivation of integrins, most likely by interfering with their binding to the extracellular matrix. This would explain the rapid detachment of GT-treated cells before they die. This mode of cell death is known as anoikis.

130

In vivo validation of GT-induced apoptosis in invasive aspergillosis

10 In vivo validation of GT-induced apoptosis in invasive aspergillosis GT is the major virulence factor that enables Aspergillus fumigatus (A.f.) to invade the lung tissue of infected patients, promoting invasive aspergillosis (IA). Previous studies suggested that the induction of apoptosis might represent a key feature of GT to mediate invasiveness226. In this study, Bak deficient mice were partially protected from IA after infection with A.f. The molecular machinery of GT-induced cell death was investigated in this thesis. GT triggered cell detachment-induced apoptosis by activating a kinase cascade, resulting in the activation of pro-apoptotic Bim. In vivo experiments were planned to address if (i) apoptosis-deficient mice were indeed protected from IA and (ii) if the pro-apoptotic RhoA/ROCK/JNK/Bim pathway was involved in the disease progression. First, the potential of A.f. (B-5233) to produce GT and subsequently induce apoptosis was tested. Increasing concentrations of A.f. conidia (1 x to 6 x 107/50 ml) were used to inoculate cell culture media. BEAS-2B cells were incubated for 6 h in the conditioned medium or with purified GT as a control (1-2 µM). Both GT and conditioned medium activated ROCK as seen by T696 phosphorylation of its substrate MYPT1 and induced gel mobility shifts of Bim (Figure 10.1). Bim phosphorylation correlated with caspase-3 cleavage.

131

In vivo validation of GT-induced apoptosis in invasive aspergillosis

Figure 10.1 | A.f. conditioned supernatant activates ROCK. BEAS-2B cells were either treated with GT (1 to 2 µM) or A.f. conditioned cell culture media (1 x to 6 x 107 conidia per 50 ml) for 6 h. The conditioned media activated ROCK as efficiently as GT as assessed by phosphorylation of its substrate MYPT at T696. Rock activity coincided with gel mobility shifts of Bim, indicating phosphorylation, and caspase-3 cleavage. Next, a protocol for the effective immunosuppression of C57BL/6 mice was established. 129/Sv mice were employed in preceding experiments in the lab, because this strain was shown to develop a profound IA after infection. C57BL/6 mice were used here, in order to study different mice that were deficient in various apoptosis regulators. All mice were immunosuppressed for four days prior to infection with A.f. (day 0) and for four days post-infection (see Figure 6.2, page 83). In a first experiment, mice were only suppressed using hydrocortisone (HC) without fungal infection. Wildtype C57BL/6 mice showed reduced white blood cell counts (WBC) and lymphocytes in peripheral blood at day 0, day 2 and day 4 (Figure 10.2). Neutrophil and monocyte counts were similar in all conditions.

132

In vivo validation of GT-induced apoptosis in invasive aspergillosis

Figure 10.2 | C57BL/6 mice are immunosuppressed by hydrocortisone (HC). C57BL/6 mice were i.p. and s.c. injected with 3 mg HC each. Immunosuppression was monitored in peripheral blood using the ADVIA 120 hematology system. WBC and lymphocyte counts were decreased in mice that received HC, compared to mice that received vehicle alone (ctrl). Neutrophil and monocyte counts were similar, but low, in all conditions.

10.1 Bak-dependent apoptosis is involved in IA C57BL/6 mice, either wild-type or deficient for Bak or Bax, were employed to establish an animal model for IA. Furthermore, these mice were used to study if apoptosis is involved in the disease progression, i.e. if they could confirm our in vitro results on GT-induced apoptosis signaling. Previous data showed that Bak-deficient, but not Baxdeficient mice did not develop IA after fungal infection, using five animals per condition226. Disease free survival of HC-injected animals was monitored after A.f. infection or PBS treatment as a control (Figure 10.3). Wild-type and Bax-/- animals were 133

In vivo validation of GT-induced apoptosis in invasive aspergillosis terminated within 10 days post infection, due to IA related symptoms. One Bak-/- and all immunosuppressed but not infected wild-type animals reached the experimental endpoint without showing any symptoms of the disease.

Figure 10.3 | Bak knock-out mice are partially protected from IA. C57BL/6 mice of the indicated genotype were immunosuppressed by HC injection and infected at day 0 with A.f. or PBS alone as a control (ctrl). Animals were inspected daily and terminated according to cancelation criteria. All wild-type and Bax deficient mice were terminated within 10 days. Two out of three Bak-/- mice, however, were terminated at later time points or after reaching the experimental endpoint. Preliminary data, no statistics applied. N = 2 for wt and ctrl, n = 3 for Bax-/- and Bak-/-. A major question in these studies was to determine the invasive potential of A.fumigatus. Therefore, lung sections were stained for fungal infiltration. We detected A.f. in the lungs of both wild-type and Bax-/- animals. However, while A.f. seems to have infiltrated surrounding tissues in the wt and Bax-/- mice, the fungal growth in Bak-/- lungs was mainly restricted to pulmonary cavities (Figure 10.4).

134

In vivo validation of GT-induced apoptosis in invasive aspergillosis

Figure 10.4 | Mice deficient of Bak, but not Bax are protected from A.f. invasion in the lung tissue. Lungs of terminated animals of the indicated treatment were inflated by injection of 1-2 ml O.C.T. Tissue TEK compound and cut in 10 µm slices. Silver staining and subsequent H&E staining were employed to visualize fungal structures (brown). In Bak-/- mice, fungal growth was mainly limited to pulmonary airways without massive infiltration of the tissue. Bax-/- and WT mice, in contrast, showed penetration of the lung epithelium by A.f. The massive fungal burden detected in histological sections might account for the partial loss of Bak-/- animals in terms of disease-free survival (Figure 10.3). Additional read-outs for A.f. invasiveness were employed to further assess the invasive potential of the fungal infection. Once invasive, A.f. infections can manifest at

135

In vivo validation of GT-induced apoptosis in invasive aspergillosis secondary sites. Fungal growth was observed in the liver of some infected animals (Figure 10.5, A). As a consequence, the livers of terminated animals were ectomized, minced through a cell strainer and cultured in RPMI. Fungal growth at the surface of the media indicated infiltration of the liver and hence the development of IA in the respective animal (Figure 10.5, B).

Figure 10.5 | Secondary infection sites of invasive A.fumigatus. Livers were ectomized after termination of infected animals to examine the systemic presence of A.f. conidia, indicative for the development of IA. (A) Fungal growth in the liver was occasionally visible, as for a wild-type mouse in this case (top right). (B) Traces of invasive fungus were amplified by cultivation of minced livers in RPMI for two to three days. Fungal invasion of the liver was confirmed by fungal growth on the surface of liver cultures. Finally, traces of A.f. were detected by nested PCR in peripheral blood. Altogether, the lung histology data, the liver culture and diagnostic PCR allowed to determine the invasiveness of A.f. for each animal (Table 10.1).

136

In vivo validation of GT-induced apoptosis in invasive aspergillosis Table 10.1 | Invasiveness of A.f. in apoptosis-deficient transgenic mice. A.f. infections were diagnosed as IA if either lung histology, liver culture or the diagnostic PCR were positive for the presence of fungal conidia outside of pulmonary cavities. Genotype

Treatment

Animals/invasive

Wild-type

PBS

2/0

Wild-type

A.F.

2/2

Bax-/-

A.F.

3/3

Bak-/-

A.F.

3/1

All wild-type and Bax-deficient mice studied in this cohort were diagnosed with IA. In contrast, only one out of three Bak-/- mice showed invasive spores. The high fungal burden in alveolar cavities might account for the premature termination of some Bak-/- animals. Although preliminary, the data indicates that A.f.-induced apoptosis via Bak might be required to invade the lung tissue of the host and to develop IA.

10.2 Establishment of Bim deficient C57BL/6 mice as an IA model GT induces apoptosis via pro-apoptotic phosphorylation of Bim. Therefore, Bim deficient C57BL/6 mice were employed to study if activation of the pro-apoptotic kinase cascade from RhoA/ROCK to Bim is involved in the development of IA in vivo. Wild-type and Bim-/- animals were treated with HC four days prior and four days post infection. WBCs were counted in peripheral blood at day 0 (the day of infection) and day 6 to monitor the immune-status of treated animals. Wild-type animals showed low levels of WBC and neutrophils at day 0 and recovered towards day 6, as the immunosuppression was relieved (Figure 10.6). Bim-/- mice showed a delayed response to the immunosuppression by HC. WBC counts were initially higher, compared to WT animals, and were reduced at day 6, two days after the last HC injection.

137

In vivo validation of GT-induced apoptosis in invasive aspergillosis

Figure 10.6 | Bim-/- mice show a delayed response to HC. C57BL/6 wild-type and Bim-/- mice were i.p and s.c. injected with 3 mg HC per injection at day -4, day -2, day 0, day 2 and day 4. Immunosuppression was monitored in peripheral blood at day 0 (d0) and day6 (d6) using the ADVIA 120 hematology system. WBC and neutrophil counts were low in wild-type mice (WT) that received HC as compared to Bim-/- mice at d0. WBC counts slightly increased at d6, two days after the last HC injection. Bim deficient mice showed a delayed response to the HC treatment, indicated by decreasing WBC counts at d6. 138

In vivo validation of GT-induced apoptosis in invasive aspergillosis Disease free survival was monitored in these animals, according to cancelation criteria. The wild-type, A.f. infected mice, showed symptoms of IA within 10 days and were terminated (Figure 10.7). Immunosuppressed but non-infected animals (ctrl), as well as fungus infected Bim-/- animals, showed no indication of fungal infection and reached the experimental endpoints.

Figure 10.7 | Bim deficient C57BL/6 mice do are protected from A.f. invasiveness. C57BL/6 mice of the indicated genotype were immunosuppressed by HC injection and infected at day 0 with A.f. or PBS alone as a control (ctrl). Animals were inspected daily and terminated according to cancelation criteria. All A.f. infected wild-type mice were terminated within 10 days. PBS treated control animals, as well as A.f. infected Bim-/animals, reached the experimental endpoint without any visible symptoms of a fungal infection. Preliminary data, no statistics applied. N = 3 for all conditions. Although these experiments are of preliminary nature, the data indicates that Bim and Bak might be involved in the invasiveness of A. fumigatus. Bim and Bak dependency on the development of IA suggests a possible role of GT-induced cell detachment and apoptosis of alveolar lung epithelial cells in tissue infiltration of the fungus and the development of IA in humans.

139

Execution of apoptosis by phosphorylated Bim

Discussion 11 Execution of apoptosis by phosphorylated Bim Previous studies in our lab employed the fungal toxin Gliotoxin (GT) as a model system to study novel pathways of apoptosis induction. These studies established that GT induces apoptosis by a JNK-mediated phosphorylation of Bim at three sites (S100, T112, S114) in a Bak-dependent manner132,226. The JNK-dependency of this process was shown by the utilization of pharmacological inhibitors and JNK1/2-deficient MEFs. Under both conditions, GT failed to induce apoptosis and the triple phosphorylation of Bim. During the course of my thesis, I contributed to this study by further characterizing the JNK-dependent phosphorylation of Bim and its significance for the interaction of Bim with anti- and proapoptotic members of the Bcl-2 family. Firstly, we needed to show that the discovered apoptotic pathway was also important for the GT-induced apoptosis of primary murine alveolar epithelial cells directly isolated from the animals (AEC Type II cells). Previous studies were mainly performed with immortalized MEFs and lung epithelial BEAS-2B cells. We could show that AEC Type II cells also underwent GT-induced apoptosis (although the GT dose needed to be increased a bit from 1 to 5 µM) in a manner that was dependent on JNKmediated Bim phosphorylation because the JNK inhibitor SP600125 prevented both GT-induced Bim phosphorylation and caspase-3 processing (Figure 8.2 and Figure 8.3). This finding confirms that the JNK/Bim signaling pathway is important for GT-induced apoptosis under a more physiological setting. Secondly, since pharmacological inhibitors such as SP600125 are not absolutely specific for the kinase described, we needed to show that Bim was a direct substrate of JNK in the GT-induced apoptosis signaling pathway. Furthermore, we needed to show that Bim was not only phosphorylated at the T112 site, as previously published, but also at two additional sites (S100 and S114). For that purpose I performed

an

in

vitro

protein

kinase

assay

with

recombinant

JNK

on

immunoprecipitated wt and triple phosphodeficient FLAG-Bim using radioactive γ32P-ATP (Figure 8.1). These results confirmed that Bim is phosphorylated at all three

140

Execution of apoptosis by phosphorylated Bim sites by JNK. Thus we provided the first direct evidence that T112 was not the only JNK phosphorylation site on Bim which increased its pro-apoptotic activity. Thirdly, it had remained elusive how the triple phosphorylation of Bim increased its pro-apoptotic potential. It was appealing to speculate that a triple phosphorylation near the BH3 domain of Bim might alter its association with pro-survival proteins or Bax and Bak. Several different mechanisms were feasible in that context, two of them are illustrated in Figure 11.1. Phosphorylation of Bim might have increased the binding affinity of Bim to Bcl-2-like survival factors thereby facilitating the release of pre-bound Bax or Bak which then would oligomerize and form a cytochrome c-releasing pore through autoactivation. This scenario focuses on the derepressor function of Bim (Figure 11.1, A). Alternatively, the triple phosphorylation might have diminished the association of Bim with inhibitory Bcl-2-like survival factors, releasing Bim to directly activate Bax/Bak. Monitoring the binding of pro-survival Bcl-2 to immunoprecipitated, FLAGtagged Bim during GT action favored the latter mode (Figure 11.1, B). GT treatment diminished Bim-Bcl-2 interactions but increased Bim-Bak interactions. Meanwhile we could show that a phosphomimetic triple mutant of Bim (S100E/T112E/S114E) also showed less interactive capacity with Mcl-1 and Bcl-xL in co-IPs (Simon Neumann and Katharina Mühlbauer, personal communication). Our finding seems to be at odds with our own report132 showing that a phosphodeficient (instead of a phosphomimetic) triple mutant of Bim showed diminished binding to Bcl-2 and Bcl-xL compared to wt Bim. At that time we performed only single time point co-immunoprecipitations (4 h GT) between wt and phosphodeficient FLAG-Bim and the survival factors. As shown in Figure 7.1 the Bim-Bcl-2 interactions however change during the course of GT action. While in the first hours triple phosphorylated Bim interacted less with Bcl-2 in co-IPs, this was not the case anymore at 4 h post GT treatment, probably due to the fact that Bim gets dephosphorylated with time. It is therefore crucial to look at the dynamics of Bcl-2 family interactions in order to make firm conclusions about their impact on apoptosis regulation.

141

Execution of apoptosis by phosphorylated Bim

Figure 11.1 | Possible mechanisms of Bim activation in response to phosphorylation. Phosphorylation of Bim at S100, T112 and S114 increases its proapoptotic activity. The underlying mechanisms remain controversial. (A) Illustration of the “derepressor model”. Bim phosphorylation was proposed to increase its binding to pro-survival Bcl-2, 4 h post GT challenge. Bim would act as a derepressor protein in that context, releasing Bax and Bak from inhibitory complexes with Bcl-2. (B) The “direct activator model”. Our data indicates that Bim interacts with pro-survival Bcl-2 in non-treated conditions. Addition of GT-induced the rapid dissociation of Bim:Bcl-2 (1-3 h). Free Bim can act as a direct activator for Bak, indicated by Bim:Bak interaction at time points were Bim was released from Bcl-2. Another mechanism by which phosphorylation of Bim may regulate its activity is through its interaction with dynein light chain 1 (DLC1). Interaction of Bim with DLC1 has been thought to inactivate Bim by sequestering it to the microtubule-associated dynein motor complex and hence restrain it from acting on mitochondria127. In this context, T112 phosphorylation was demonstrated to release Bim from DLC1 to induce its apoptotic activity238. This view has changed with recent work showing that Bim is always localized to mitochondria via its C-terminal membrane anchor and therefore does not necessarily have to interact with microtubules to be kept in check128.

142

Execution of apoptosis by phosphorylated Bim However, the binding of DLC1 with Bim has been confirmed and interestingly the DLC binding motif (T110/Q111) overlaps with the region that is triple phosphorylated by JNK (T112/S114) (Figure 11.2). Therefore the question is how Bim phosphorylation affects Bim-DLC1 binding and what consequence does this have on the pro-apoptotic activity of Bim. Recently, Bim was shown to form homodimers or oligomers of higher order in a DLC1 dependent way in healthy cells (Daniel Frank and Georg Häcker, personal communication). These data are in line with previous studies reporting that Bim is present in high molecular weight complexes239. Bim mutants that lack the DLC1 binding motif failed to oligomerize and these mutants seemed to be more apoptotic and to bind less strongly to Bcl-2-like survival factors as wt Bim (Prafull Kumar and Georg Häcker, personal communication). This is in line with our data using the triple phosphomimetic mutant, and triple phosphorylated Bim in response to GT in cellulo. Given the close proximity of the DLC binding site and the phosphorylated T112/S114 region, it is therefore reasonable to speculate that in healthy cells Bim is inactivated at mitochondria by DLC1-dependent formation of high molecular weight complexes. Phosphorylation of Bim at T112/S114, for example by JNK, would then release Bim from these DLC1 containing complexes (Figure 11.2) in line with earlier reports130,238. Consequently, Bim phosphorylation would diminish its oligomerization capacity and facilitate its direct interaction with and activation of Bax and Bak to promote apoptosis. Interestingly, we have recently found in our lab that the triple phosphomimetic Bim mutant S110E/T112E/S114E forms smaller complexes on blue native gels than wt Bim (Simon Neumann and Katharina Mühlbauer, personal communication) supporting the notion that triple phosphorylation of Bim breaks its high order oligomerization. More studies using various Bim mutants and cellular systems are needed to confirm the role of the triple phosphorylation of Bim by JNK for apoptosis induction. This also includes mouse models where phosphodeficient and –mimetic Bim mutants are knocked into the Bim gene locus by CRISPR/Cas technology and then testing the role of the triple phosphorylation for cell homeostasis, for example in the immune system where Bim is absolutely crucial.

143

Deciphering the JNK activating signaling cascade

Figure 11.2 | DLC1 sequesters Bim. DLC1 induces the formation of Bim dimers, oligomers and high molecular weight complexes (Daniel Frank and Georg Häcker, personal communication). The residues T110 and Q111 are essential for this interaction. The close proximity of the pro-apoptotic phosphorylation sites might explain the release of Bim from DLC1 to induce apoptosis. Domain structure was drawn using the Illustrator for Biological Sequences (IBS V1.0).

12 Deciphering the JNK activating signaling cascade The stress response kinase JNK has been shown to be involved in both pro- and antiapoptotic cell fates depending on the stimulus (reviewed by Liu and Lin153). Growth factor withdrawal and UV irradiation are the most prominent stimuli that engage proapoptotic JNK signaling240. We showed that GT employs JNK to execute apoptosis via triple phosphorylation of Bim. The major question of this thesis was to identify the GT induced signaling events that result in the activation of JNK. JNK belongs to the family of mitogen activated protein kinases (MAPK) and is therefore activated by a hierarchical MAP kinase module of the form MAPKKkinase

MAPKkinase

MAPK.

Therefore, it was obvious that JNK must be activated by upstream MAP-like kinases.

12.1 GSK3 enhances the apoptotic response to GT Glycogen synthase kinase-3 (GSK-3) is a protein kinase that was initially described to regulate metabolism by inactivating glycogen synthase. Today it is also known to regulate cell survival and death by direct or indirect regulation of Bcl-2 proteins. Interestingly, GSK3 was described to induce Bim-dependent cell death in concert with JNK241. Although this study reported that mutual phosphorylation of c-Jun regulates

144

Deciphering the JNK activating signaling cascade Bim on a transcriptional level, we wanted to investigate a possible involvement of GSK3 in the transmission of GT signaling upstream or in concert with JNK. Inhibition of GSK3 by Axon-1126, a small-molecule inhibitor that exhibits high effectiveness and specificity towards GSK3242, rescued cells from GT-induced caspase activation (Figure 9.5). However, neither the activating phosphorylation of Jnk nor the pro-apoptotic phosphorylation of Bim were reduced by this inhibitor. Thus, active GSK3 does not seem to act upstream of the JNK/Bim signaling pathway to contribute to GT-induced apoptosis, but must enhance apoptosis by another, JNK/Bim-independent mechanism. Interestingly, GSK3 was shown to phosphorylate the acetyltransferase TIP60 which subsequently acetylates p53 at Lysine120 to induce Puma144. However, Puma induction by GSK3 was not detected in response to GT on mRNA or protein level (data not shown). Moreover, Puma does not seem to be required for GT-induced apoptosis132. This further substantiated our previous findings that GT mainly employs the BH3-only protein Bim to execute apoptosis. Thus, GSK3 rather acts on a pathway that does not directly activate Bim but modulates the pro-apoptotic activity of Bim, for example by regulating its antagonists, the Bcl-2 survival factors. It is known that active GSK3 phosphorylates the anti-apoptotic Bcl-2 family member Mcl-1 to induce its proteasomal degradation by ubiquitylation146,243. Inhibition of GSK3 therefore stabilized Mcl-1 in response to GT (Figure 9.7). Increased levels of Mcl-1 may then sequester activated Bim and hence protect the cells from GT-induced apoptosis. Interestingly, the GSK3 phosphorylation that marks Mcl-1 for degradation (S140 in mice) requires a priming phosphorylation by JNK (T144)146. In that sense GSK3 and JNK cooperate to regulate GT-induced apoptosis. The exact mechanisms by which GSK3 is activated by GT remain to be investigated. GSK3 activation most likely requires preceding PI3K and AKT inactivation. This may occur due to the inactivation of FAK by GT as PI3K and AKT are downstream targets of FAK signaling. Phosphorylation deficient mutants of Mcl-1 that cannot be targeted for degradation should be employed in future experiments to further elucidate the protective role of Mcl-1 stabilization in this context. In addition, we need to understand why triple phosphorylated Bim is still effectively sequestered to stabilized Mcl-1 (one mechanism of cell death protection by GSK3 inhibition) but less so to Bcl-2 as shown in Figure 7.1. Recent evidence suggests that Bim more likely interacts with Mcl-1 than

145

Deciphering the JNK activating signaling cascade with other Bcl-2 survival factors (Georg Häcker, personal communication) indicating that elevated levels of Mcl-1 may be more protective against GT-induced apoptosis driven by JNK-mediated Bim phosphorylation than elevated levels of Bcl-2 or Bcl-xL. If true, then the GSK3-mediated degradation of Mcl-1 to release Bim combined with a further activation of Bim by triple phosphorylation might be an optimal strategy of GT to maximally activate Bax/Bak-mediated MOMP.

12.2 Gliotoxin employs RhoA to trigger a JNK activating kinase cascade As GSK3 was not the upstream protein kinase, which activated JNK on the GT apoptosis signaling pathway, we had to look for another kinase that performed this job. The two most prominent MAPKKs for JNK are MKK4 (JNKK1) and MKK7 (JNKK2)153,244,245. MKK4 was shown to phosphorylate both p38 and JNK, whereas MKK7 is specific for JNK. These two MAPKKs relay numerous signals to JNK, including the regulation of differentiation and proliferation as well as apoptosis. The signaling outcome is context-specific and depends on the formation of signaling complexes including different MAPKKKs, MKK4/7 and JNK246. The crucial role of MKK4 and MKK7 in the activation of JNK by GT was unveiled by employing MEFs deficient of these two kinases. Single and double knock-outs showed reduced annexin-V staining and caspase activity in response to GT treatment (Figure 9.3). Interestingly, the double knock-out cells showed the highest degree of protection indicating that both MAPKKs are involved in the activation of JNK. Indeed, both MKKs were activated by GT because they were both phosphorylated at their activating residues (pS257/pT261 for MKK4 and pS271/pT275 for MKK7) as evidenced by immunoblotting with phosphospecific antibodies (Figure 9.4, A). Consistent with these results, MEFs deficient for both MKKs abolished the activating phosphorylation of JNK, as well as the pro-apoptotic phosphorylation of Bim, in response to GT (Figure 9.4, B). Data obtained by genetic deletion of MKKs should be confirmed in future experiments by using inhibitors against MKK4 and/or MKK7. Inhibition of MKK7 could be achieved by overexpressing the GADD45β protein, which has been shown to be a potent inhibitor of MKK7247. In addition, small-molecule inhibitors have been reported, like trihydroxyisoflavone for MKK4 or HWY336, targeting both MKK4 and MKK7248,249.

146

Deciphering the JNK activating signaling cascade However, as for other protein kinase inhibitors the specificity and efficiency of these inhibitors is questionable, also due to their rare usage. Specific inhibitors would help to further evaluate the individual contribution of MKK4 versus MKK7 in the transmission of GT-induced signaling. It seems reasonable that both MKKs are somehow involved since full activation of JNK requires the phosphorylation at T183 and at Y185. MKK4 has a preferential affinity for Y185, whereas MKK7 preferentially phosphorylates T183244,246,250. The phosphorylation of T183 is sufficient to induce basal activity of Jnk, which can be fine-tuned by additional phosphorylation at Y185. It is therefore not surprising, that both kinases cooperate to activate JNK in the response to GT. Final clarification of the individual MKK4/7 influence on JNK activation could be gained by re-expressing the individual proteins in MKK4/7 double knock-out MEFs. We tried to achieve this, but had been unsuccessful so far maybe due to a toxic effect of one or the other kinase when largely overexpressed. Nevertheless, the data presented here strongly supports the hypothesis that both MKK4 and MKK7 transmit the apoptotic GT signal to JNK, which subsequently phosphorylates Bim. The upstream protein that activates MKK4 and MKK7 is likely to be another protein kinase. Fourteen MAPKKKs have been reported to activate MKK4 and MKK7. These include apoptosis signal-regulating kinase1 (ASK1), mixed lineage kinase 1-4 (MLK), TAK1246 and ROCK1 and 2. We focused on ROCK as a candidate upstream kinase for the following reasons: GT-treated adherent cells detached within 1 h, implying a role for cytoskeleton regulating proteins. The key regulators of actin dynamics are Rho associated kinase (ROCK) and the mammalian homologue of Drosophila diaphanous (mDIA). mDIA regulates actin polymerization. However, neither actin polymerization nor de-polymerization could be observed in response to GT (Figure 9.8). Instead we could show that GT induced the contraction of stress fibers during the detachment process pointing towards a possible involvement of ROCK in this process (Figure 9.8). Interestingly ROCK has hitherto not been associated with the classical MAPKKK tier for MKK4 and MKK7 in a pro-apoptotic context. As discussed in chapter 2.3.1, ROCK was previously proposed to be involved in JNK-mediated apoptotic signaling because it activated JNK and c-Jun phosphorylation during arsenic oxide induced cell death of chronic myelogenous leukemia (CML) cell lines251. Furthermore, ROCK was shown to regulate cell migration via Jnk and cJun252,253. This data indicated a possible link between ROCK and JNK.

147

Deciphering the JNK activating signaling cascade By using two pharmacological inhibitors, H-1152 and Y-27632, and cells in which the expression of ROCK1 was knocked-down by shRNA, we could show that ROCK was required for GT-induced MKK4/7, JNK and Bim phosphorylation, and subsequent caspase-3 activation and hence apoptosis (Figure 9.11, 9.12 and 9.13). The blockage of apoptosis under these conditions was as efficient as after treatment with the general caspase inhibitor QVD (Figure 9.10). Moreover, ROCK was indeed activated in response to GT as its substrate myosin binding subunit of myosin phosphatase 1 (MYPT1) was phosphorylated with a similar kinetic as MKK4 and JNK phosphorylation/activation (Figure 9.17). We used two different ROCK inhibitors to avoid the risk of data misinterpretation due to off-target effects. When the specificity of the inhibitors was tested against a panel of 70 protein kinases, it was found that they inhibited protein kinase C-related protein kinase (PRK2) with similar efficiency as ROCK242. Their specificity towards the described MAPKKKs for MKK-JNK was not tested in this study. Therefore, an off-target effect on other MAPKKKs cannot be excluded at this point. However, both ROCK inhibitors performed very similar in terms of their protective potential against GTinduced apoptosis, indicating that the observed rescue was due to inhibition of ROCK. Furthermore, although RNAi approaches are also prone to elicit off-target effects, the combined data with two inhibitors and the efficient knock-down of ROCK1 strongly argues that ROCK was essential to transmit the apoptotic GT signal to MKK4 and JNK (Figure 9.14 and 9.15) In vitro kinase assays should, however, still be performed in order to further show if ROCK directly phosphorylates MKK4/7 or if the observed phosphorylation is via an intermediate protein kinase. Most strikingly, ROCK inhibition or knock-down of ROCK1 not only prevented the induction of cell death but also prevented the GT-induced cell detachment. This notion further substantiates the critical role of ROCK at the crossroad between regulation of actin dynamics and cell death signaling. The Rho family of small GTPases were the obviously suspected upstream activators of ROCK in the GT-induced apoptosis signaling pathway because GTP-bound Rho is known to activate ROCK, by binding to its C-terminal Rho-binding domain158. Indeed, RhoA was activated within 40 min of GT treatment (Figure 9.16) and inhibition of Rho by the bacterial toxin C3 diminished ROCK activation (Figure 9.17) as well as MKK4, JNK and Bim phosphorylation (Figure 9.19). Inhibitors of

148

Deciphering the JNK activating signaling cascade CDC42 and Rac had no effect on GT signaling. Moreover, as seen with ROCK inhibition, C3 maintained the epithelial morphology of GT-treated cells (Figure 9.18), similar to inhibition of ROCK. The tools to study Rho activation by small-molecules are limited, compared to the broad spectrum of kinase inhibitors. The Rho family, including Rho, CDC42 and Rac, are highly interconnected and regulated by positive as well as negative feedback loops. Knock-down of these proteins should therefore be avoided, because the interpretations of data might not be reliable and/or difficult to reproduce (personal communication, Dr. Schwan). Therefore we used two bacterial toxins in order to modify Rho activity. The C3 toxin inhibits RhoA by ADP-ribosylation, preventing interaction with GEFs and increasing the binding to GDIs, thereby generating an inactive, cytosolic pool of RhoA. Cytotoxic necrotizing factor y (CNFy) in contrast, activates RhoA by deamidation of a catalytic glutamine in the GTPase domain, leading to inhibition of GTP hydrolysis254,255. These toxins allowed to study the GT-induced signaling by immunoblotting, however as the name (cytotoxic necrotizing factor y) implies, these toxins can induce necrosis. Cell death assays were therefore omitted in this chapter. Other read-outs like ROCK activity by phosphorylation of MYPT1 at T696 were used instead. Another interesting technique to study Rho activity is the utilization of RhoFRET sensors. A variety of constructs have been designed that undergo FRET when Rho is activated256. The application of such constructs in future experiments will allow for a better spatio-temporal resolution of Rho activation at a subcellular scale. With our analysis we clearly established a novel anoikis pathway induced by GT-induced cell detachment. This pathway activates RhoA which in turn stimulates ROCK activity that is then used to activate MKK4/7 and JNK. JNK-mediated the triple phosphorylation of Bim at S100, T112 and S114 to stimulate its pro-apoptotic activity towards Bak-mediated MOMP and apoptosis (Figure 12.1). In addition, ROCK regulates actin dynamics and contractility by modulating myosin light chain (MLC) ATPase activity and inhibiting MLC phosphatase, necessary for cell detachment.

149

The role of integrins and FAK in the anoikis signaling induced by GT

Figure 12.1 | The proposed GT-induced kinase cascade. We suggest that GT triggers cell death by activation of ROCK in a RhoA dependent fashion. Active ROCK functions as a MAPKKK for JNK, phosphorylating MKK4 and MKK7. These MAPKKs consequently phosphorylate JNK. JNK then executes apoptosis by phosphorylating Bim and hence increasing its pro-apoptotic potential.

13 The role of integrins and FAK in the anoikis signaling induced by GT GT rapidly induces cell detachment before apoptotic morphological alterations can be seen (Figure 9.22). This suggests that perturbance of integrins and its associated focal adhesion complex might be involved in the apoptosis signaling process. Focal adhesion kinase (FAK) is a crucial kinase in this complex, transmitting signals from integrins to cellular targets, including MAPKs. FAK is recruited to the tail of integrins by the adaptor protein paxillin. We found, by using life cell microscopy, that in response to GT GFP-paxillin translocates from a focal adhesion pattern to vesicles (Figure 9.23) which moved from the focal plane at the plasma membrane to the cytosol (data not shown). These vesiclelike structures seem to consist of early endosomes as co-transfection of GFP-paxillin

150

The role of integrins and FAK in the anoikis signaling induced by GT with mRuby-Endo14, an endosomal marker, showed a partial localization of paxillin at endosomal membranes (Figure 9.24). Although the nature of this interaction and the vesicular distribution of paxillin remains speculative, it helps to explain the observed degradation of Paxillin in response to GT (Figure 9.25). Paxillin may be degraded by fusion of early endosomes with lysosomes. Inhibitors of lysosomal degradation like chloroquine should be employed in future experiments to further elucidate this mechanism. Nevertheless, our data showed that GT somehow targeted focal adhesions, leading to an endocytosis of components of the complex such as paxillin. This notion is further underlined by the rapid loss of the FAK activating phosphorylation at Y397 (for details see chapter 2.3.2) by GT treatment (Figure 9.26). This phosphorylation is a read-out for the attachment of lung epithelial cells and MEFs to the extracellular matrix, since untreated adhesion cells showed phosphorylation of FAK at this site (Figure 9.26). Importantly, FAK inactivation starts 30 min after GT treatment which coincides with the activation of RhoA (40 min). Therefore it was logical to speculate that FAK inactivation is linked to RhoA activation. Holinstat et al. first established the link between FAK and RhoA178. They showed that active FAK suppressed RhoA in the context of endothelial barrier integrity by phosphorylating and activating p190RhoGAP, a GTPase of Rho proteins. Consistent with their findings, we observed that after FAK inactivation by GT, p190RhoGAP was dephosphorylated and inactivated, thereby allowing RhoA activation and subsequent JNK activation via ROCK (Figure 9.27). On the contrary, overexpression of FAK could at

least

partially

suppress

GT-induced

apoptosis

via

diminished

JNK

phosphorylation/activation. The problem was that GT was still capable of inactivating retrovirally transduced, overexpressed FAK. This explains why the overexpression only delayed but not entirely inhibited the apoptotic response to GT (Figure 9.30). Strikingly, even mutated FAK that was constitutively recruited to the membrane by myristoylation and hence activated, was efficiently inactivated by GT (data not shown). Alternatively, a kinase active mutant of FAK should be employed to prevent the GT-induced activation of RhoA. In this respect, we have recently started a collaboration with the lab of Prof. Winfried Weber (BIOSS, University of Freiburg) to use a light inducible construct of FAK which allows the reversible activation of FAK by irradiation with blue light in the timescale of one minute (W. Weber, unpublished).

151

The role of integrins and FAK in the anoikis signaling induced by GT Importantly, inhibition of FAK by pharmacological inhibitors, without the addition of GT, induced cell death and Rho activation (Figure 9.28) along with inactivation of FAK/p190RhoGAP and activation of the ROCK dependent kinase cascade (Figure 9.29). These findings indicate that GT induces RhoA activation via the dephosphorylation and inactivation of FAK which then can no longer inactivate RhoA by p190RhoGAP (Figure 13.1).

Figure 13.1 | FAK inactivation triggers RhoA activity. FAK stimulates the GAP activity of p190RhoGAP (p190) to suppress RhoA in adherent cells. Cell detachment inactivates FAK and p190, which results in the activation of RhoA to trigger the proapoptotic kinase cascade via ROCK. In that sense, the inhibition of the GTPase activating protein p190 would be sufficient to activate RhoA in adherent cells. However, it is possible that Rho activation is also positively regulated by Rho-GEFs. Such GEFs have not yet been identified in this study. A possible tool to do so is the inactive RhoA mutant T19N. It was shown that GEFs can bind to this mutant but cannot activate it. A GST-fusion protein GST-RhoA T19N coupled to agarose beads can then pull down candidate GEFs from whole lysates257. We applied this method to lysates of GT-treated cells but could not detect any specific Rho GEFs in our pull downs (by silver staining on SDS-PAGE, data not shown). Yet, the participation of specific GEFs cannot be excluded at this point. Interesting targets that have been reported to act downstream of FAK include ARHGEF28 (p190RhoGEF), ARHGEF11 (PDZ-RhoGEF), leukemia-associated RhoGEF (LARG, ARHGEF12) and ARHGEF1 (p115RhoGEF)258. The induction of apoptosis competent signaling from RhoA seems to be dependent on the activation of RhoA by FAK. Interestingly, cells treated with the

152

The role of integrins and FAK in the anoikis signaling induced by GT bacterial toxin CNFy, which activates RhoA, did not show apoptotic signaling (Figure 9.19, A). A possible explanation for this differential response of RhoA might be that the focal adhesion complex provides a scaffold to link RhoA and ROCK activation to MKK4/7 by close proximity. Therefore, activation of RhoA at focal adhesions would enable it to transduce pro-apoptotic signaling, whereas induction of RhoA by other stimuli such as CNFy rather promotes its function to regulate the cytoskeleton. The involvement of focal adhesion signaling in GT-induced apoptosis encouraged us to speculate that integrins are targeted by GT in order to inactivate FAK and detach cells. Integrins represent an appealing target for GT toxicity for the following reasons: (i) GT is suspected to act on proteins at the plasma membrane. Once it enters the cell, it is reduced and inactivated by glutathione. (ii) GT was proposed to exhibit its toxic effects by binding to cysteines of target proteins. Indeed, integrins contain up to 56 cysteines, including a conserved cysteine-rich region for activation169. In addition, integrins depend on disulfide bonds that induce and stabilize the active conformation and help to position the head domain for ligand binding168,259. We found for the first time that GT is able to break disulfide bonds and thereby may interfere with integrin binding to the ECM. In situ incubation of recombinant human insulin with GT induced the reduction of insulin chains that are linked by disulfide bonds (Appendix 1). This data additionally raised the possibility that GT can act as a crosslinking agent, because silver staining showed bands at higher molecular weight than insulin. It is therefore possible that GT alters integrin activity by interfering with the network of disulfide bridges. Mass spectrometry data confirmed this notion by showing covalent modification of Cys158 of recombinant αV and Cys258 of integrin β3 by GT in vitro. These cysteines are located in the ligand binding domain of integrin αVβ3. Both cysteines are conserved among their respective subunit and form disulfide bridges to other cysteines (Figure 9.32). The conservation of these cysteines among all integrin alpha and beta species indicates that GT may not only covalently modify them in αVβ3 but also in other integrin chains. The Cys258 of the β3 subunit is located in the b-I domain, responsible for ligand binding. Additionally, it is involved in the formation of a disulfide bridge with Cys299166. The impact of this residue on ligand binding and integrin activity is so far not further characterized, but it might contribute to the positioning of the head domain or the function of the metal ion dependent adhesion site (MIDAS).

153

The role of integrins and FAK in the anoikis signaling induced by GT The integrin αV Cys158 is located in an exposed loop of the second repeat of the seven bladed β propeller (see Figure 2.12). This propeller represents the ligand binding domain of alpha chains. The ligand binding specificity of the propeller is mainly determined in the second and third repeat260–262. Swapping of a small region of this exposed loop (Asp154-Ala159) from αVβ1 with the region from α5β1, consequently changed the ligand binding properties of αVβ1 to those of α5β1262. These data indicate an essential role of Cys158 in the selectivity for integrin ligands and the ability to engage the ECM. It is therefore tempting to speculate that the GT modification of this residue has an impact on the ability to engage ligands and hence mediate cell adhesion. In contrast to the ligand binding domains, modification of the conserved cysteine-rich region for integrin activation could not be observed by MS. This region is located in the I-EGF domains 1 to 4 of the β subunits (see chapter 2.3.2). Peptide mapping revealed that no peptides were obtained after tryptic digest of this region, although in silico digest indicated the presence of trypsin cleavage sites (data not shown). Modified conditions and/or a different protease should be used in order to yield peptides from this region for further evaluation of GT binding. Together, these in situ data suggest that GT may interfere with the integrin binding to extracellular matrix components and hence trigger cell detachment-induced apoptosis. Indeed, inactivation of integrins by GT was confirmed in a more physiological setup by staining integrin activity on BEAS-2B cells with a fluorescently-labeled RGD peptide derived from fibronectin (Figure 9.35) which binds to a variety of integrin heterodimers including αVβ3, αIIbβ3, αVβ6, αVβ1, α5β1, α8β1, α4β1 and α4β7263. Although it was difficult to say which integrins were labeled, it became evident that the overall integrin activity was reduced within 30 min after GT treatment which perfectly coincided with the kinetic of FAK inactivation (30 min) and subsequent RhoA activation (40 min) (Figure 9.35). Most strikingly, our idea that GT targeted integrins to promote cell detachment and apoptosis was supported by the fact that suspension cells were protected from GT-induced cytotoxicity (Figure 9.33). These cells showed no apoptotic response to GT and failed to activate the ROCK-dependent kinase cascade, although they expressed integrins on their cell surface (Appendix 2). The expressed integrins exhibited low RGD peptide binding when kept in suspension (Figure 9.37). Re-

154

The role of integrins and FAK in the anoikis signaling induced by GT activation of integrins by attaching suspension cells to fibronectin-coated plates restored the ability of these cells to phosphorylate MKK4 and JNK and to activate caspases and apoptosis in response to GT (Figure 9.38). This dependency of GTinduced cell death on the presence of active integrins is further emphasized by early studies showing that GT can inhibit bacterial growth but fails to kill bacteria207. Based on all these data, we propose the following mechanism of GT-induced apoptosis (Figure 13.2): GT inactivates integrins by covalently modifying cysteines in the ligand binding domain of the alpha (Cys158) and the beta (Cys258) subunits. This results in the detachment of cells because integrins cannot effectively interact with extracellular components any more. Cell detachment is sensed by the focal adhesion complex, located at the cytosolic tail of integrins, in a way that FAK is inactivated and paxillin is endocytosed and degraded. FAK can no longer stimulate p190RhoGAP which usually keeps RhoA activity at low levels in healthy, attached cells. Therefore RhoA gets activated and stimulates ROCK, which in turn triggers a pro-apoptotic kinase cascade via MKK4/7 and JNK resulting in the triple phosphorylation of Bim to execute GT-induced cell death. This pathway can be considered as the first molecularly described, physiologically relevant signaling pathway for anoikis (greek for “homelessness”). It is stimulated by GT but can be mimicked by FAK inhibition using selective FAK inhibitors. How paxillin is degraded and how this contributes to the activation of RhoA remains to be determined. It is speculative that Paxillin degradation inactivates FAK by inducing the release from focal adhesions, thereby inducing a cytosolic, auto-inhibited pool of FAK that can be re-activated by recruitment to the membrane. The claim that integrin inactivation triggers the described GT signaling could be further evaluated using antagonistic antibodies that are known to inactivate integrins264.

155

The role of integrins and FAK in the anoikis signaling induced by GT

Figure 13.2 | Proposed mechanism of GT-induced apoptosis. Integrins engage the ECM in healthy adherent cells, recruiting paxillin and focal adhesion kinase (FAK). FAK suppresses RhoA activity by phosphorylating and hence activating p190RhoGAP (p190). GT inactivates integrins to detach cells. Cell detachment induces paxillin degradation and inactivation of FAK. p190 is no longer able to suppress Rho activity, leading to the stimulation of ROCK and induction of the pro-apoptotic kinase cascade.

156

The role of integrins and FAK in the anoikis signaling induced by GT

13.1 GT as a model to study anoikis The ability of cells to escape cell detachment-induced apoptosis (anoikis) is a wellestablished hallmark of cancer cells. Accurate execution of anoikis prevents cell growth at ectopic sites by inducing apoptosis of cells that detach from their environment. Anoikis therefore represents a critical mechanism to prevent metastasis and tumor migration171. The pro-survival signaling induced by integrins in adherent cells via FAK is well characterized and described in chapter 2.3.2. Cell death is inhibited in these cells by activation of MAPKs and PI3K signaling. Activation of ERK and PI3K/AKT downstream of focal adhesions keeps Bim at low levels as described before (chapter 2.3.1). The molecular understanding of the induction and execution of anoikis, however, is poorly understood. The identification of anoikis inducing pathways is hampered by the lack of physiological systems to study this mode of cell death. The most common model is the coating of cell culture plates with Poly-Hema (PH) or similar polymers. In these type of studies, cells are detached by trypsin and subsequently cultured on PH plates, forcing the cells to stay in suspension and thereby undergo anoikis. This approach allows the comparison of adherent cells and suspension cells (of the same cell type) and the molecular changes attributed to the respective context172,182–184,265. The PH studies suggested that anoikis is due to inactivation of ERK and AKT signaling. Because ERK usually triggers the proteasomal degradation of Bim through phosphorylation in healthy cells, Bim gets stabilized during anoikis and activates Bax/Bak-mediated MOMP171,265,266. Indeed, inactivation of ERK and AKT can be seen in PH cultured BEAS-2B cells (Appendix 3). By contrast, GT-induced anoikis showed activation instead of inactivation of ERK. The reason for this difference is currently unknown. Maybe GT-induced cell detachment induces pro-survival signaling in an attempt to rescue the cell. Prolonged exposure to the stress factor overcomes this response and induces cell death. Interestingly, ERK phosphorylation was highest in cells immediately after trypsinization and declined over time (Appendix 3), arguing in the same direction. The discrepancy between PH- and GT-induced anoikis may be explained by the fundamental difference of these two systems. Trypsinization of cells and cultivation on PH describes processes that are related to the inhibition of cell adhesion. The employment of GT, however, allows the identification of cellular signals

157

The role of integrins and FAK in the anoikis signaling induced by GT that are directly triggered by the detachment of cells. Cell detachment by GT might therefore represent a more physiological system to study anoikis than trypsinization followed by cultivation of the cells on PH coated plates. The stabilization of Bim alone, as suggested by the PH system of anoikis, might not be sufficient to induce apoptosis as stabilized Bim can still be inactivated by prosurvival Bcl-2 family members. It is therefore reasonable to speculate that an additional trigger that enhances the pro-apoptotic activity of Bim activity is required to execute anoikis. Employing GT as a model system of anoikis, revealed that this direct activation of Bim is mediated by inactivation of FAK and subsequent activation of a RhoA/ROCK dependent kinase cascade leading to the triple phosphorylation of Bim (Figure 13.2). Bim was released from Bcl-2 in response to GT to execute Bax/Bak-mediated MOMP (Figure 7.1). The notion that phosphorylation deficient Bim mutants failed to induce GT-induced anoikis underlines that the RhoA-Jnk-Bim pathway is a more important regulator of anoikis than simple Bim stabilization132. The suggestion that GT could represent a functional anoikis inducer is enhanced by our finding that PI3K/GSK3 signaling is involved in this process (see chapter 12.1). Inhibition of GSK3 rescued GT induced cell death, most likely by stabilization of Mcl-1. This might explain the protective effect of FAK-dependent PI3K activation to suppress anoikis175, because PI3K inhibits GSK3 via AKT-dependent phosphorylation. Furthermore, the hypothesis that anoikis is RhoA-dependent is in line with previous findings. Microarray analysis of TGF-β transformed normal murine mammary gland cells showed upregulation of the miRNA miR-155, which targets RhoA267. TGF-β induced EMT could be blocked by overexpressing RhoA indicating that RhoA has an essential role in preventing oncogenic cell migration and invasion. Although cell death was not monitored in this study, it is plausible that the oncogenic effect is mediated by resistance to anoikis due to loss of RhoA171. A better understanding of the molecular events that govern anoikis might help to characterize the mechanisms employed by cancer cells to evade this cell death program. GT is not only an interesting target to study anoikis but, to our best knowledge, the first small molecule described that is possibly able to directly inactivate integrins. Such a molecule could be an important tool to further investigate integrin function, focal adhesion signaling and cytoskeletal changes during cell detachment.

158

Apoptosis-deficient mice as a model for invasive aspergillosis (IA) Finally, it is important to mention that integrins are most likely not the only target of GT. Early studies described GT as an inhibitor of NFκB with immunosuppressive functions218–220, whereas other groups reported that GT can induce the secretion of IL-8222. Most likely, GT has a broad spectrum of possible targets that are preferentially located in or at the plasma membrane and require cysteines for proper function. Nevertheless, the fast kinetic of GT-induced cell detachment might allow to ignore the above-mentioned side effects.

14 Apoptosis-deficient mice as a model for invasive aspergillosis (IA) Our in vitro data clearly showed that GT triggers cell detachment and subsequent apoptosis via a JNK/Bim-dependent signaling pathway in immortalized human lung epithelial cells (BEAS-2B) as well as in primary murine alveolar epithelial cells (AEC type II). The same signaling pathway is activated by incubating these cells with the supernatant of cultured Aspergillus fumigatus (A.f.). It is therefore very likely that after inhalation A.f. may perform a similar pro-apoptotic action on the lung epithelium in vivo and that the major mediator of this action would be GT. This would explain why the epithelial barrier is broken and the fungus becomes invasive. However, since the physiology of the lung is quite complex, in vivo experiments are needed to evaluate the impact of the above described pathway on the disease progression in mice. Importantly, all the subsequent discussion points have to be considered on a preliminary basis since we have not yet been able to reproduce the in vivo data for a statistical analysis. Bim, Bax or Bak deficient mice were studied in order to evaluate the contribution of apoptosis in general (Bax-/- and Bak-/-), or the influence of the described anoikis pathway (Bim-/-) on the disease progression. The previously reported 129/Sv mouse strain could not be used, because the described knock-out animals originated from a C57BL/6 background. An effective immunosuppression was crucial to achieve effective pulmonary infection with A.fumigatus. C57BL/6 showed a robust reduction of white blood cells in response to hydrocortisone injection (Figure 10.2) while maintaining levels of neutrophils, as expected268. Non-neutropenic mice models were previously shown to be more susceptible to develop IA as compared to neutropenic

159

Apoptosis-deficient mice as a model for invasive aspergillosis (IA) models224. Therefore, hydrocortisone was employed since this immunosuppressant is known to maintain neutrophil levels. Two out of three Bak deficient mice showed a prolonged survival after infection compared to wild-type or Bax-/- mice. These data were in line with a previous report that GT-induced apoptosis in vitro and IA after A.f. infection in vivo predominantly required Bak not Bax226. These data suggest that apoptosis may be involved in compromising the lung epithelial barrier integrity during A.f. infections. Consistent with this notion the A.f. fungus did not invade into the lung and surrounding tissues and did not seem to spread to other organs such as the liver in Bak-/- mice. Instead, the lung sections of these mice showed that pulmonary airways were filled with fungus (Figure 10.4). It is therefore not surprising that these mice were only partially protected in terms of survival. Although apoptosis deficiency may prevent lung infiltration, it does not prohibit fungal growth in the lumen of the alveoli. If then the artificially high fungal load cannot be cleared by the recovering immune system, the mice may get problems with breathing and the proper function of their lungs. It is therefore important to control the invasiveness of the infection in these mice. In addition, lower concentrations of conidia might help to amplify the observed protection. We found that the ELISA method, which detects the A.f. surface antigen galactomannan was more reliable than the PCR reaction measuring fungal DNA. The ELISA test can be applied on BAL fluids sampled from the lung as well as on peripheral blood. Therefore, it allows the confirmation of a successful lung infection as well as an invasive spreading to other organs (through the blood). In addition, invasiveness can be shown by staining fungal structures on lung sections as well as by detecting fungal growth in distant organs such as the liver. The sensitivity of the latter approach is however quite low and it is prone to huge variability because a high fungal load is necessary to yield fungal growth in the liver. Taken together, these assays allow a good estimation of the invasiveness of A.f. and hence the influence of the genetic model used in the experiment. The most promising data was obtained from our in vivo studies derived from Bim-/- mice. Although only three mice have yet been analyzed, it was surprising to see that all of them survived the challenge with GT. In this context, it is however important to mention that these mice showed less immunosuppression. Bim was reported to be the major BH3-only protein involved in mediating glucocorticoid-induced cell death of

160

Apoptosis-deficient mice as a model for invasive aspergillosis (IA) immune cells269–271. Indeed, two out of three animals showed higher levels of white blood cells as compared to wild-type animals at day 0 (Figure 10.6). This insufficient immunosuppression might be the reason for the observed protection of the animals. Remaining immune activity could have prevented the conversion of A.f. spores into hyphens, which is necessary to produce GT and breach the lung epithelial barrier. One of these animals, however, showed full immunosuppression comparable to wild-type animals and still survived the treatment. It is important to note that Bim mice showed similar WBC at day 6 as wild-type animals at day 0. This indicates that immunosuppression by hydrocortisone might be delayed in Bim-/- mice. In this case, the infection scheme (Figure 6.2) should simply be delayed and/or prolonged in order to circumvent the described obstacle in immunosuppression. In any case, we need to repeat our Bim-/- experiments with higher mouse numbers to obtain statistical significance. Moreover, we consider to use conditional mice in the future which could be induced by the Cre-lox system to delete Bim only in lung epithelial cells and not in the hematopoietic system. Alternatively, we could cross Bim-/- mice with Rag1-deficient mice to circumvent the need for hydrocortisone-induced immunosuppression. These mice lack the recombination-activating gene (Rag) and hence do not possess mature B and T lymphocytes, but still produce normal numbers of neutrophils272. Finally, we could envisage to transfer healthy bone marrow into non-lethally irradiated Bim-/- mice (adoptive transfer) to repopulate these mice with wt immune cells. These mice could then be normally immunosuppressed but would have lung epithelial cells that are deficient for Bim expression.

14.1 Inhibition of apoptosis, a possible treatment for IA? The scope of this thesis was to identify the molecular signaling pathways employed by GT and A.f. to induce anoikis. Characterization of such signaling cascades is needed in order to identify possible drug targets to block the invasive potential of A.f. infections. The strategy in this context is not to target the fungus directly but to prevent a systemic infection of the host. The current antifungal drugs fail to improve the devastating mortality rates associated with IA, mainly due to their pharmacokinetics, crossreaction with corticosteroids and resistance mechanisms of the fungus. Restraining the fungus in the lung of the patient would increase the time to apply different treatments, to

161

Apoptosis-deficient mice as a model for invasive aspergillosis (IA) restore the patient’s immune system or to remove fungal bulbs by surgery. Targeting the ability of GT to induce apoptosis might therefore help to prevent invasion of the host. Preliminary data presented here and by others indicates that apoptosis is involved in the invasive potential of A.f. in vivo 226. GT-induced anoikis could be blocked pharmacologically or genetically by interfering with the apoptosis signaling pathway on several levels. Inhibition of ROCK represents the most appealing target in that context. This strategy does not only prevented apoptosis, but also maintained the epithelial morphology of BEAS-2B cells by inhibiting cell detachment. This reflects the dual role of Rho and ROCK in executing apoptosis and regulating cell shape in the context of GT. In terms of barrier integrity, it would be most promising to address targets that are upstream of MKK4/7 in order to regulate cell death and cell morphology. It is therefore tempting to speculate that inhibition of ROCK might help to maintain epithelial barrier integrity by blocking cell detachment and apoptosis in humans. This hypothesis is encouraged by reports that show a protective effect of RhoA inhibition by GDIs on endothelial barrier integrity in vivo273–275. These studies used Rho-GDI-1 deficient mice to show the importance of RhoA in that context. Pharmacological inhibition of ROCK might resemble this effect by protecting epithelial barriers. Furthermore, inhibition of ROCK was recently shown to prevent the uptake of conidia in epithelial cells by regulating cofilin276. These findings support the role of ROCK, although the uptake of conidia by epithelial cells is most likely not involved in the disease progression198. The relevance of ROCK as a potential drug target to treat patients is further encouraged by the existence of ROCK inhibitors like Fasudil. Fasudil (HA-1077) has already been approved in Japan for the treatment of cerebral vasospasm. However, the high diversity of ROCK targets, in addition to MKK4/7, increases the likelihood of adverse effects of an anti-ROCK therapy. In addition, Fasudil was shown to have only a moderate selectivity towards ROCK and inhibits a number of other kinases as well277,278. The in vivo administration of Fasudil to treat IA is therefore questionable. Over 170 ROCK inhibitors with increased selectivity have been generated during the last years, however, their in vivo application potential needs to be evaluated yet. The identification of a ROCK inhibitor with low off-target effects, and

162

Apoptosis-deficient mice as a model for invasive aspergillosis (IA) hence an acceptable spectrum of adverse effects, might be beneficial in the treatment of IA. Targeting the GT toxin directly may be the most promising and most effective strategy to block the invasion of A.f. in lung tissue. A respective patent has been put forward by our lab to pharmacologically inhibit GT by a compound named GTinhibit. If successful, such a compound will inhibit all GT-induced signaling pathways in lung epithelial cells, including those, which are not yet known. This approach mimics Aspergillus strains that lack GT production. These strains, lacking the GliP gene in the GT biosynthesis cluster, were shown to be non pathogenic212. Targeting the toxin directly avoids alterations of cellular signaling events and it is thus expected to exhibit the least possible adverse side effects. Details about the GTinhibit compound cannot be discussed here before it will be tested in vivo.

163

Bibliography

Bibliography 1.

Sulston, J. E. Post-embryonic development in the ventral cord of Caenorhabditis elegans. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 275, 287–97 (1976).

2.

Melino, G. The Sirens’ song. Nature 412, 23 (2001).

3.

Kroemer, G. et al. Classification of cell death: recommendations of the Nomenclature Committee on Cell Death. Cell Death Differ. 12 Suppl 2, 1463–7 (2005).

4.

Galluzzi,

L.

et

al.

Molecular

definitions

of

cell

death

subroutines:

recommendations of the Nomenclature Committee on Cell Death 2012. Cell Death Differ. 19, 107–20 (2012). 5.

Fulda, S. & Kögel, D. Cell death by autophagy: emerging molecular mechanisms and implications for cancer therapy. Oncogene 34, 5105–13 (2015).

6.

Hotchkiss, R. S., Strasser, A., McDunn, J. E. & Swanson, P. E. Cell Death. N Engl J Med 361, 1570–83 (2009).

7.

Lotze, M. T. & Tracey, K. J. High-mobility group box 1 protein (HMGB1): nuclear weapon in the immune arsenal. Nat. Rev. Immunol. 5, 331–42 (2005).

8.

Holler, N. et al. Fas triggers an alternative, caspase-8-independent cell death pathway using the kinase RIP as effector molecule. Nat. Immunol. 1, 489–95 (2000).

9.

Degterev, A. et al. Chemical inhibitor of nonapoptotic cell death with therapeutic potential for ischemic brain injury. Nat. Chem. Biol. 1, 112–9 (2005).

10.

Degterev, A. et al. Identification of RIP1 kinase as a specific cellular target of necrostatins. Nat. Chem. Biol. 4, 313–21 (2008).

11.

Pasparakis, M. & Vandenabeele, P. Necroptosis and its role in inflammation. Nature 517, 311–320 (2015).

12.

Zhang, D.-W. et al. RIP3, an energy metabolism regulator that switches TNFinduced cell death from apoptosis to necrosis. Science 325, 332–6 (2009).

164

Bibliography 13.

Cho, Y. S. et al. Phosphorylation-driven assembly of the RIP1-RIP3 complex regulates programmed necrosis and virus-induced inflammation. Cell 137, 1112–23 (2009).

14.

He, S. et al. Receptor interacting protein kinase-3 determines cellular necrotic response to TNF-alpha. Cell 137, 1100–11 (2009).

15.

Vandenabeele, P., Declercq, W., Van Herreweghe, F. & Vanden Berghe, T. The role of the kinases RIP1 and RIP3 in TNF-induced necrosis. Sci. Signal. 3, re4 (2010).

16.

Li, J. et al. The RIP1/RIP3 necrosome forms a functional amyloid signaling complex required for programmed necrosis. Cell 150, 339–50 (2012).

17.

Sun, L. et al. Mixed Lineage Kinase Domain-like Protein Mediates Necrosis Signaling Downstream of RIP3 Kinase. Cell 148, 213–227 (2012).

18.

Zhao, J. et al. Mixed lineage kinase domain-like is a key receptor interacting protein 3 downstream component of TNF-induced necrosis. Proc. Natl. Acad. Sci. U. S. A. 109, 5322–7 (2012).

19.

Remijsen, Q. et al. Depletion of RIPK3 or MLKL blocks TNF-driven necroptosis and switches towards a delayed RIPK1 kinase-dependent apoptosis. Cell Death Dis 5, e1004 (2014).

20.

Jouan-Lanhouet, S. et al. Necroptosis, in vivo detection in experimental disease models. Semin. Cell Dev. Biol. 35, 2–13 (2014).

21.

Cai, Z. et al. Plasma membrane translocation of trimerized MLKL protein is required for TNF-induced necroptosis. Nat. Cell Biol. 16, 55–65 (2014).

22.

Chen, X. et al. Translocation of mixed lineage kinase domain-like protein to plasma membrane leads to necrotic cell death. Cell Res. 24, 105–21 (2014).

23.

Dondelinger, Y. et al. MLKL Compromises Plasma Membrane Integrity by Binding to Phosphatidylinositol Phosphates. Cell Rep. 7, 971–981 (2014).

24.

Vandenabeele, P., Galluzzi, L., Vanden Berghe, T. & Kroemer, G. Molecular mechanisms of necroptosis: an ordered cellular explosion. Nat. Rev. Mol. Cell Biol. 11, 700–14 (2010).

25.

Silke, J. & Brink, R. Regulation of TNFRSF and innate immune signalling 165

Bibliography complexes by TRAFs and cIAPs. Cell Death Differ. 17, 35–45 (2010). 26.

Micheau, O. & Tschopp, J. Induction of TNF receptor I-mediated apoptosis via two sequential signaling complexes. Cell 114, 181–90 (2003).

27.

Kaiser, W. J. et al. RIP3 mediates the embryonic lethality of caspase-8-deficient mice. Nature 471, 368–72 (2011).

28.

Zhang, H. et al. Functional complementation between FADD and RIP1 in embryos and lymphocytes. Nature 471, 373–6 (2011).

29.

Upton, J. W., Kaiser, W. J. & Mocarski, E. S. Virus inhibition of RIP3-dependent necrosis. Cell Host Microbe 7, 302–13 (2010).

30.

Seifert, L. et al. The necrosome promotes pancreatic oncogenesis via CXCL1 and Mincle-induced immune suppression. Nature 532, 245–249 (2016).

31.

Dixon, S. J. et al. Ferroptosis: an iron-dependent form of nonapoptotic cell death. Cell 149, 1060–72 (2012).

32.

Yang, W. S. & Stockwell, B. R. Synthetic lethal screening identifies compounds activating iron-dependent, nonapoptotic cell death in oncogenic-RAS-harboring cancer cells. Chem. Biol. 15, 234–45 (2008).

33.

Yang, W. S. & Stockwell, B. R. Ferroptosis: Death by Lipid Peroxidation. Trends Cell Biol. 26, 165–176 (2016).

34.

Yang, W. S. et al. Regulation of ferroptotic cancer cell death by GPX4. Cell 156, 317–31 (2014).

35.

Warner, G. J. et al. Inhibition of selenoprotein synthesis by selenocysteine tRNA[Ser]Sec lacking isopentenyladenosine. J. Biol. Chem. 275, 28110–9 (2000).

36.

Dolma, S., Lessnick, S. L., Hahn, W. C. & Stockwell, B. R. Identification of genotype-selective antitumor agents using synthetic lethal chemical screening in engineered human tumor cells. Cancer Cell 3, 285–296 (2003).

37.

Zhou, W. & Yuan, J. Necroptosis in health and diseases. Semin. Cell Dev. Biol. 35, 14–23 (2014).

38.

Sadagurski, M. et al. IRS2 increases mitochondrial dysfunction and oxidative

166

Bibliography stress in a mouse model of Huntington disease. J. Clin. Invest. 121, 4070–81 (2011). 39.

Jiang, L. et al. Ferroptosis as a p53-mediated activity during tumour suppression. Nature 520, 57–62 (2015).

40.

Green, D. R. & Llambi, F. Cell Death Signaling. Cold Spring Harb. Perspect. Biol. 7, a006080– (2015).

41.

Levine, B. & Yuan, J. Autophagy in cell death: an innocent convict? J. Clin. Invest. 115, 2679–88 (2005).

42.

Galluzzi, L. et al. Essential versus accessory aspects of cell death: recommendations of the NCCD 2015. Cell Death Differ. 22, 58–73 (2015).

43.

Balakrishnan, K., Wierda, W. G., Keating, M. J. & Gandhi, V. Gossypol, a BH3 mimetic, induces apoptosis in chronic lymphocytic leukemia cells. Blood 112, 1971–80 (2008).

44.

Meng, Y. et al. Natural BH3 mimetic (-)-gossypol chemosensitizes human prostate cancer via Bcl-xL inhibition accompanied by increase of Puma and Noxa. Mol. Cancer Ther. 7, 2192–202 (2008).

45.

Voss, V. et al. The pan-Bcl-2 inhibitor (-)-gossypol triggers autophagic cell death in malignant glioma. Mol. Cancer Res. 8, 1002–16 (2010).

46.

Lian, J., Karnak, D. & Xu, L. A natural BH3 mimetic induces autophagy in apoptosis-resistant prostate cancer via modulating Bcl-2–Beclin1 interaction at endoplasmic reticulum. Autophagy 6, 1201–3 (2010).

47.

Nakahira, K. & Choi, A. M. K. Autophagy: a potential therapeutic target in lung diseases. Am. J. Physiol. Lung Cell. Mol. Physiol. 305, L93–107 (2013).

48.

Kerr, J. F., Wyllie, A. H. & Currie, A. R. Apoptosis: a basic biological phenomenon with wide-ranging implications in tissue kinetics. Br. J. Cancer 26, 239–57 (1972).

49.

Brenner, S. The genetics of Caenorhabditis elegans. Genetics 77, 71–94 (1974).

50.

Ellis, H. M. & Horvitz, H. R. Genetic control of programmed cell death in the nematode C. elegans. Cell 44, 817–29 (1986).

167

Bibliography 51.

Hengartner, M. O. & Horvitz, H. R. C. elegans cell survival gene ced-9 encodes a functional homolog of the mammalian proto-oncogene bcl-2. Cell 76, 665–76 (1994).

52.

Fuchs, Y. & Steller, H. Programmed cell death in animal development and disease. Cell 147, 742–58 (2011).

53.

Lindsten, T. et al. The combined functions of proapoptotic Bcl-2 family members bak and bax are essential for normal development of multiple tissues. Mol. Cell 6, 1389–99 (2000).

54.

Jacobson, M. D., Weil, M. & Raff, M. C. Programmed Cell Death in Animal Development. Cell 88, 347–354 (1997).

55.

Barres, B. A. & Raff, M. C. Axonal control of oligodendrocyte development. J. Cell Biol. 147, 1123–8 (1999).

56.

Opferman, J. T. & Korsmeyer, S. J. Apoptosis in the development and maintenance of the immune system. Nat. Immunol. 4, 410–5 (2003).

57.

Hanahan, D. & Weinberg, R. A. Hallmarks of cancer: the next generation. Cell 144, 646–74 (2011).

58.

Vila,

M.

&

Przedborski,

S.

Targeting

programmed

cell

death

in

neurodegenerative diseases. Nat. Rev. Neurosci. 4, 365–75 (2003). 59.

Elmore, S. Apoptosis: a review of programmed cell death. Toxicol. Pathol. 35, 495–516 (2007).

60.

Ravichandran, K. S. Find-me and eat-me signals in apoptotic cell clearance: progress and conundrums. J. Exp. Med. 207, 1807–17 (2010).

61.

Savill, J. & Fadok, V. Corpse clearance defines the meaning of cell death. Nature 407, 784–8 (2000).

62.

Thornberry, N. A. & Lazebnik, Y. Caspases: enemies within. Science 281, 1312– 6 (1998).

63.

Stennicke, H. R. & Salvesen, G. S. Properties of the caspases. Biochim. Biophys. Acta 1387, 17–31 (1998).

64.

Taylor, R. C., Cullen, S. P. & Martin, S. J. Apoptosis: controlled demolition at the

168

Bibliography cellular level. Nat. Rev. Mol. Cell Biol. 9, 231–241 (2008). 65.

Tait, S. W. & Green, D. R. Mitochondria and cell death: outer membrane permeabilization and beyond. Nat. Rev. cell Biol. 11, 621–632 (2010).

66.

Boatright, K. M. et al. A unified model for apical caspase activation. Mol. Cell 11, 529–41 (2003).

67.

Oberst, A. et al. Inducible dimerization and inducible cleavage reveal a requirement for both processes in caspase-8 activation. J. Biol. Chem. 285, 16632–42 (2010).

68.

Degterev, A., Boyce, M. & Yuan, J. A decade of caspases. Oncogene 22, 8543– 8567 (2003).

69.

Earnshaw, W. C., Martins, L. M. & Kaufmann, S. H. Mammalian Caspases: Structure , Activation, Substrates and functions during apoptosis. Annu. Rev. Biochem. 68, 383–424 (1999).

70.

Thiede, B., Treumann, A., Kretschmer, A., Söhlke, J. & Rudel, T. Shotgun proteome analysis of protein cleavage in apoptotic cells. Proteomics 5, 2123– 2130 (2005).

71.

Croft, D. R. et al. Actin-myosin–based contraction is responsible for apoptotic nuclear disintegration. J. Cell Biol. 168, 245–255 (2005).

72.

Lüthi, A. U. & Martin, S. J. The CASBAH: a searchable database of caspase substrates. Cell Death Differ. 14, 641–650 (2007).

73.

Liu, X., Zou, H., Slaughter, C. & Wang, X. DFF, a heterodimeric protein that functions downstream of caspase-3 to trigger DNA fragmentation during apoptosis. Cell 89, 175–84 (1997).

74.

Krammer, P. H., Arnold, R. & Lavrik, I. N. Life and death in peripheral T cells. Nat. Rev. Immunol. 7, 532–542 (2007).

75.

Kischkel, F. C. et al. Cytotoxicity-dependent APO-1 (Fas/CD95)-associated proteins form a death-inducing signaling complex (DISC) with the receptor. EMBO J. 14, 5579–88 (1995).

76.

Chang, D. W., Xing, Z., Capacio, V. L., Peter, M. E. & Yang, X. Interdimer processing mechanism of procaspase-8 activation. EMBO J. 22, 4132–42 169

Bibliography (2003). 77.

Muzio, M. An Induced Proximity Model for Caspase-8 Activation. J. Biol. Chem. 273, 2926–2930 (1998).

78.

Lavrik, I. N. et al. Analysis of CD95 threshold signaling: triggering of CD95 (FAS/APO-1) at low concentrations primarily results in survival signaling. J. Biol. Chem. 282, 13664–71 (2007).

79.

Legembre, P. et al. Induction of apoptosis and activation of NF-kappaB by CD95 require different signalling thresholds. EMBO Rep. 5, 1084–9 (2004).

80.

Schleich, K. et al. Stoichiometry of the CD95 death-inducing signaling complex: experimental and modeling evidence for a death effector domain chain model. Mol. Cell 47, 306–19 (2012).

81.

Dickens, L. S. et al. A death effector domain chain DISC model reveals a crucial role for caspase-8 chain assembly in mediating apoptotic cell death. Mol. Cell 47, 291–305 (2012).

82.

Riley, J. S., Malik, A., Holohan, C. & Longley, D. B. DED or alive: assembly and regulation of the death effector domain complexes. Cell Death Dis. 6, e1866 (2015).

83.

Majkut, J. et al. Differential affinity of FLIP and procaspase 8 for FADD’s DED binding surfaces regulates DISC assembly. Nat. Commun. 5, 3350 (2014).

84.

Liu, X., Kim, C. N., Yang, J., Jemmerson, R. & Wang, X. Induction of apoptotic program in cell-free extracts: requirement for dATP and cytochrome c. Cell 86, 147–57 (1996).

85.

Scorrano, L. & Korsmeyer, S. J. Mechanisms of cytochrome c release by proapoptotic BCL-2 family members. Biochem. Biophys. Res. Commun. 304, 437–44 (2003).

86.

Riedl, S. J. et al. Structural basis for the inhibition of caspase-3 by XIAP. Cell 104, 791–800 (2001).

87.

Riedl, S. J. & Salvesen, G. S. The apoptosome: signalling platform of cell death. Nat. Rev. Mol. Cell Biol. 8, 405–13 (2007).

88.

Acehan, D. et al. Three-dimensional structure of the apoptosome: implications 170

Bibliography for assembly, procaspase-9 binding, and activation. Mol. Cell 9, 423–32 (2002). 89.

Zou, H., Henzel, W. J., Liu, X., Lutschg, A. & Wang, X. Apaf-1, a human protein homologous to C. elegans CED-4, participates in cytochrome c-dependent activation of caspase-3. Cell 90, 405–13 (1997).

90.

Yu, X. et al. A structure of the human apoptosome at 12.8 A resolution provides insights into this cell death platform. Structure 13, 1725–35 (2005).

91.

Renatus, M., Stennicke, H. R., Scott, F. L., Liddington, R. C. & Salvesen, G. S. Dimer formation drives the activation of the cell death protease caspase 9. Proc. Natl. Acad. Sci. U. S. A. 98, 14250–5 (2001).

92.

Chipuk, J. E., Bouchier-Hayes, L. & Green, D. R. Mitochondrial outer membrane permeabilization during apoptosis: the innocent bystander scenario. Cell Death Differ. 13, 1396–402 (2006).

93.

Tsujimoto, Y., Cossman, J., Jaffe, E. & Croce, C. M. Involvement of the bcl-2 gene in human follicular lymphoma. Science 228, 1440–3 (1985).

94.

Vaux, D. L., Cory, S. & Adams, J. M. Bcl-2 gene promotes haemopoietic cell survival and cooperates with c-myc to immortalize pre-B cells. Nature 335, 440– 2 (1988).

95.

Moldoveanu, T., Follis, A. V., Kriwacki, R. W. & Green, D. R. Many players in BCL-2 family affairs. Trends Biochem. Sci. 39, 101–111 (2014).

96.

Llambi, F. et al. A Unified Model of Mammalian BCL-2 Protein Family Interactions at the Mitochondria. Mol. Cell 44, 517–531 (2011).

97.

Youle, R. J. & Strasser, A. The BCL-2 protein family: opposing activities that mediate cell death. Nat. Rev. Mol. Cell Biol. 9, 47–59 (2008).

98.

Chipuk, J. E., Moldoveanu, T., Llambi, F., Parsons, M. J. & Green, D. R. The BCL2 family reunion. Moll Cell. 37, 299–310 (2010).

99.

Petros, A. M., Olejniczak, E. T. & Fesik, S. W. Structural biology of the Bcl-2 family of proteins. Biochim. Biophys. Acta 1644, 83–94 (2004).

100. Muchmore, S. W. et al. X-ray and NMR structure of human Bcl-xL, an inhibitor of programmed cell death. Nature 381, 335–41 (1996).

171

Bibliography 101. Gavathiotis, E. et al. BAX activation is initiated at a novel interaction site. Nature 455, 1076–81 (2008). 102. Gavathiotis, E., Reyna, D. E., Davis, M. L., Bird, G. H. & Walensky, L. D. BH3triggered structural reorganization drives the activation of proapoptotic BAX. Mol. Cell 40, 481–92 (2010). 103. Willis, S. N. et al. Apoptosis Initiated When BH3 Ligands Engage Multiple Bcl-2 Homologs, Not Bax or Bak. Science (80-. ). 315, 856–859 (2007). 104. Willis, S. N. et al. Proapoptotic Bak is sequestered by Mcl-1 and Bcl-xL, but not Bcl-2, until displaced by BH3-only proteins. Genes Dev. 19, 1294–305 (2005). 105. Letai, A. et al. Distinct BH3 domains either sensitize or activate mitochondrial apoptosis, serving as prototype cancer therapeutics. Cancer Cell 2, 183–92 (2002). 106. Wei, M. C. et al. tBID, a membrane-targeted death ligand, oligomerizes BAK to release cytochrome c. Genes Dev. 14, 2060–71 (2000). 107. Leber, B., Lin, J. & Andrews, D. W. Embedded together: the life and death consequences of interaction of the Bcl-2 family with membranes. Apoptosis 12, 897–911 (2007). 108. Delbridge, A. R. D. & Strasser, A. The BCL-2 protein family, BH3-mimetics and cancer therapy. Cell Death Differ. 22, 1071–80 (2015). 109. Strasser, A., Cory, S. & Adams, J. M. Deciphering the rules of programmed cell death to improve therapy of cancer and other diseases. EMBO J. 30, 3667–83 (2011). 110. Oltersdorf, T. et al. An inhibitor of Bcl-2 family proteins induces regression of solid tumours. Nature 435, 677–81 (2005). 111. Tse, C. et al. ABT-263: a potent and orally bioavailable Bcl-2 family inhibitor. Cancer Res. 68, 3421–8 (2008). 112. van Delft, M. F. et al. The BH3 mimetic ABT-737 targets selective Bcl-2 proteins and efficiently induces apoptosis via Bak/Bax if Mcl-1 is neutralized. Cancer Cell 10, 389–99 (2006). 113. Nakano, K. & Vousden, K. H. PUMA, a novel proapoptotic gene, is induced by 172

Bibliography p53. Mol. Cell 7, 683–94 (2001). 114. Oda, E. et al. Noxa, a BH3-only member of the Bcl-2 family and candidate mediator of p53-induced apoptosis. Science 288, 1053–8 (2000). 115. Lima, R. T. et al. MicroRNA regulation of core apoptosis pathways in cancer. Eur. J. Cancer 47, 163–74 (2011). 116. del Peso, L., González-García, M., Page, C., Herrera, R. & Nuñez, G. Interleukin3-induced phosphorylation of BAD through the protein kinase Akt. Science 278, 687–9 (1997). 117. O’Connor, L. et al. Bim: a novel member of the Bcl-2 family that promotes apoptosis. EMBO J. 17, 384–95 (1998). 118. BCL2L11.

at

119. Bouillet, P. et al. Proapoptotic Bcl-2 relative Bim required for certain apoptotic responses, leukocyte homeostasis, and to preclude autoimmunity. Science 286, 1735–8 (1999). 120. Piñon, J. D., Labi, V., Egle, a & Villunger, a. Bim and Bmf in tissue homeostasis and malignant disease. Oncogene 27, S41–S52 (2008). 121. Dijkers, P. F., Medema, R. H., Lammers, J. W., Koenderman, L. & Coffer, P. J. Expression of the pro-apoptotic Bcl-2 family member Bim is regulated by the forkhead transcription factor FKHR-L1. Curr. Biol. 10, 1201–4 (2000). 122. Puthalakath, H. et al. ER stress triggers apoptosis by activating BH3-only protein Bim. Cell 129, 1337–49 (2007). 123. Ley, R., Balmanno, K., Hadfield, K., Weston, C. & Cook, S. J. Activation of the ERK1/2 Signaling Pathway Promotes Phosphorylation and Proteasomedependent Degradation of the BH3-only Protein, Bim. J. Biol. Chem. 278, 18811–18816 (2003). 124. Luciano, F. et al. Phosphorylation of Bim-EL by Erk1/2 on serine 69 promotes its degradation via the proteasome pathway and regulates its proapoptotic function. Oncogene 22, 6785–6793 (2003). 125. Ewings, K. E. et al. ERK1/2-dependent phosphorylation of BimEL promotes its 173

Bibliography rapid dissociation from Mcl-1 and Bcl-xL. EMBO J. 26, 2856–2867 (2007). 126. Harada, H., Quearry, B., Ruiz-vela, A. & Korsmeyer, S. J. Survival factor-induced extracellular signal-regulated kinase phosphorylates BIM , inhibiting its association with BAX and proapoptotic activity. Proc. Natl. Acad. Sci. 101, 15313–7 (2004). 127. Puthalakath, H., Huang, D. C., O’Reilly, L. A., King, S. M. & Strasser, A. The proapoptotic activity of the Bcl-2 family member Bim is regulated by interaction with the dynein motor complex. Mol. Cell 3, 287–96 (1999). 128. Wilfling, F. et al. BH3-only proteins are tail-anchored in the outer mitochondrial membrane and can initiate the activation of Bax. Cell Death Differ. 19, 1328–36 (2012). 129. Balmanno, K. & Cook, S. J. Tumour cell survival signalling by the ERK1/2 pathway. Cell Death Differ. 16, 368–77 (2009). 130. Ley, R., Ewings, K. E., Hadfield, K. & Cook, S. J. Regulatory phosphorylation of Bim: sorting out the ERK from the JNK. Cell Death Differ. 12, 1008–1014 (2005). 131. Hübner, A., Barret, T., Flavell, R. A. & Davis, R. J. Multi-site Phosphorylation Regulates Bim Stability and Apoptotic Activity. Moll Cell. 30, 415–425 (2008). 132. Geissler, A. et al. Apoptosis induced by the fungal pathogen gliotoxin requires a triple phosphorylation of Bim by JNK. Cell Death Differ. 20, 1317–1329 (2013). 133. Wei, M. C. et al. Proapoptotic BAX and BAK: a requisite gateway to mitochondrial dysfunction and death. Science 292, 727–30 (2001). 134. Czabotar, P. E., Lessene, G., Strasser, A. & Adams, J. M. Control of apoptosis by the BCL-2 protein family: implications for physiology and therapy. Nat. Rev. Mol. Cell Biol. 15, 49–63 (2014). 135. Edlich, F. et al. Bcl-x(L) retrotranslocates Bax from the mitochondria into the cytosol. Cell 145, 104–16 (2011). 136. Todt, F. et al. Differential retrotranslocation of mitochondrial Bax and Bak. EMBO J 34, 67–80 (2015). 137. Czabotar, P. E. et al. Bax Crystal Structures Reveal How BH3 Domains Activate Bax and Nucleate Its Oligomerization to Induce Apoptosis. Cell 152, 519–531 174

Bibliography (2013). 138. Brouwer, J. M. et al. Bak Core and Latch Domains Separate during Activation, and Freed Core Domains Form Symmetric Homodimers. Mol. Cell 55, 938–946 (2014). 139. Peng, R. et al. Targeting Bax interaction sites reveals that only homooligomerization sites are essential for its activation. Cell Death Differ. 20, 744– 54 (2013). 140. Dewson, G. et al. To trigger apoptosis, Bak exposes its BH3 domain and homodimerizes via BH3:groove interactions. Mol. Cell 30, 369–80 (2008). 141. Salvador-Gallego, R. et al. Bax assembly into rings and arcs in apoptotic mitochondria is linked to membrane pores. EMBO J 35, 389–401 (2016). 142. Llambi, F. et al. Article BOK Is a Non-canonical BCL-2 Family Effector of Apoptosis Regulated by ER-Associated Degradation Article BOK Is a Noncanonical BCL-2 Family Effector of Apoptosis Regulated by ER-Associated Degradation. Cell 165, 421–33 (2016). 143. Zhang, X., Tang, N., Hadden, T. J. & Rishi, A. K. Akt, FoxO and regulation of apoptosis. Biochim. Biophys. Acta - Mol. Cell Res. 1813, 1978–1986 (2011). 144. Charvet, C. et al. Phosphorylation of Tip60 by GSK-3 Determines the Induction of PUMA and Apoptosis by p53. Mol. Cell 42, 584–596 (2011). 145. Maurer, U., Charvet, C., Wagman, A. S., Dejardin, E. & Green, D. R. Glycogen synthase kinase-3 regulates mitochondrial outer membrane permeabilization and apoptosis by destabilization of MCL-1. Mol. Cell 21, 749–60 (2006). 146. Maurer, U., Preiss, F., Brauns-Schubert, P., Schlicher, L. & Charvet, C. GSK-3 at the crossroads of cell death and survival. J. Cell Sci. 127, 1369–78 (2014). 147. Wada, T. & Penninger, J. M. Mitogen-activated protein kinases in apoptosis regulation. Oncogene 23, 2838–2849 (2004). 148. Tanaka, N. et al. Differential involvement of p38 mitogen-activated protein kinase kinases MKK3 and MKK6 in T-cell apoptosis. EMBO Rep. 3, 785–91 (2002). 149. Dhanasekaran, D. N. & Reddy, E. P. JNK signaling in apoptosis. Oncogene 27, 6245–6251 (2008). 175

Bibliography 150. Dérijard, B. et al. JNK1: a protein kinase stimulated by UV light and Ha-Ras that binds and phosphorylates the c-Jun activation domain. Cell 76, 1025–37 (1994). 151. Fan, M., Goodwin, M. E., Birrer, M. J. & Chambers, T. C. The c-Jun NH(2)terminal protein kinase/AP-1 pathway is required for efficient apoptosis induced by vinblastine. Cancer Res. 61, 4450–8 (2001). 152. Behrens, A., Sibilia, M. & Wagner, E. F. Amino-terminal phosphorylation of cJun regulates stress-induced apoptosis and cellular proliferation. Nat. Genet. 21, 326–9 (1999). 153. Liu, J. & Lin, A. Role of JNK activation in apoptosis: a double-edged sword. Cell Res. 15, 36–42 (2005). 154. Donovan, N., Becker, E. B. E., Konishi, Y. & Bonni, A. JNK phosphorylation and activation of BAD couples the stress-activated signaling pathway to the cell death machinery. J. Biol. Chem. 277, 40944–9 (2002). 155. Wang, X. T. et al. Opposing effects of Bad phosphorylation at two distinct sites by Akt1 and JNK1/2 on ischemic brain injury. Cell. Signal. 19, 1844–1856 (2007). 156. Yamamoto, K., Ichijo, H. & Korsmeyer, S. J. BCL-2 is phosphorylated and inactivated by an ASK1/Jun N-terminal protein kinase pathway normally activated at G(2)/M. Mol. Cell. Biol. 19, 8469–78 (1999). 157. Kharbanda, S. et al. Translocation of SAPK/JNK to mitochondria and interaction with Bcl-x(L) in response to DNA damage. J. Biol. Chem. 275, 322–7 (2000). 158. Etienne-Manneville, S. & Hall, A. Rho GTPases in cell biology. Nature 420, 629– 635 (2002). 159. Street, C. A. & Bryan, B. A. Rho Kinase Proteins—Pleiotropic Modulators of Cell Survival and Apoptosis. Anticancer Res. 31, 3645–3657 (2011). 160. Rebillard, A. et al. Cisplatin-induced apoptosis involves a Fas-ROCK-ezrindependent actin remodelling in human colon cancer cells. Eur. J. Cancer 46, 1445–55 (2010). 161. Hebert, M. et al. Rho-ROCK-Dependent Ezrin-Radixin-Moesin Phosphorylation Regulates Fas-Mediated Apoptosis in Jurkat Cells. J. Immunol. 181, 5963–5973 (2008).

176

Bibliography 162. Sebbagh, M. et al. Caspase-3-mediated cleavage of ROCK I induces MLC phosphorylation and apoptotic membrane blebbing. Nat. Cell Biol. 3, 346–52 (2001). 163. Coleman & Olson. Rho GTPase signalling pathways in the morphological changes associated with apoptosis. Cell Death Differ. 9, 493–504 (2002). 164. Boudreau, N. J. & Jones, P. L. Extracellular matrix and integrin signalling: the shape of things to come. Biochem. J. 339, 481–488 (1999). 165. Anthis, N. J. & Campbell, I. D. The tail of integrin activation. Trends Biochem Sci 36, 191–198 (2011). 166. Calvete, J. J., Henschen, A. & González-Rodríguez, J. Assignment of disulphide bonds in human platelet GPIIIa. A disulphide pattern for the beta-subunits of the integrin family. Biochem. J. 274, 63–71 (1991). 167. Kashiwagi, H. et al. A mutation in the extracellular cysteine-rich repeat region of the beta3 subunit activates integrins alphaIIbbeta3 and alphaVbeta3. Blood 93, 2559–68 (1999). 168. Yan, B. & Smith, J. W. Mechanism of Integrin Activation by Disulfide Bond Reduction. Biochemistry 40, 8861–8867 (2001). 169. Beglova, N., Blacklow, S. C., Takagi, J. & Springer, T. A. Cysteine-rich module structure reveals a fulcrum for integrin rearrangement upon activation. Nat Struct Mol Biol 9, 282–287 (2002). 170. Kamata, T. et al. Critical cysteine residues for regulation of integrin alphaIIbbeta3 are clustered in the epidermal growth factor domains of the beta3 subunit. Biochem. J. 378, 1079–1082 (2004). 171. Paoli, P., Giannoni, E. & Chiarugi, P. Anoikis molecular pathways and its role in cancer progression. Biochim. Biophys. Acta - Mol. Cell Res. 1833, 3481–3498 (2013). 172. Le Gall, M. et al. The p42/p44 MAP kinase pathway prevents apoptosis induced by anchorage and serum removal. Mol. Biol. Cell 11, 1103–1112 (2000). 173. Calalb, M. B., Polte, T. R. & Hanks, S. K. Tyrosine phosphorylation of focal adhesion kinase at sites in the catalytic domain regulates kinase activity: a role

177

Bibliography for Src family kinases. Mol. Cell. Biol. 15, 954–963 (1995). 174. Lietha, D. et al. Structural Basis for the Autoinhibition of Focal Adhesion Kinase. Cell 129, 1177–1187 (2007). 175. Chen, H.-C., Appeddu, P. A., Isoda, H. & Guan, J.-L. Phosphorylation of Tyrosine 397 in Focal Adhesion Kinase Is Required for Binding Phosphatidylinositol 3Kinase. J. Biol. Chem. 271, 26329–26334 (1996). 176. Schlaepfer, D. D., Hanks, S. K., Hunter, T. & Geer, P. van der. Integrin-mediated signal transduction linked to Ras pathway by GRB2 binding to focal adhesion kinase. 372, 786–791 (1994). 177. Mitra, S. K., Hanson, D. a. & Schlaepfer, D. D. Focal adhesion kinase: in command and control of cell motility. Nat. Rev. Mol. Cell Biol. 6, 56–68 (2005). 178. Holinstat, M. et al. Suppression of RhoA Activity by Focal Adhesion Kinaseinduced Activation of p190RhoGAP: Role in regulation of Endothelial Permeability. J. Biol. Chem. 281, 2296–2305 (2006). 179. Masiero, L., Lapidos, K. a, Ambudkar, I. & Kohn, E. C. Regulation of the RhoA pathway in human endothelial cell spreading on type IV collagen: role of calcium influx. J. Cell Sci. 112 ( Pt 1, 3205–13 (1999). 180. Noren, N. K., Arthur, W. T. & Burridge, K. Cadherin engagement inhibits RhoA via p190RhoGAP. J. Biol. Chem. 278, 13615–8 (2003). 181. Bass, M. D. et al. p190RhoGAP is the convergence point of adhesion signals from alpha 5 beta 1 integrin and syndecan-4. J. Cell Biol. 181, 1013–26 (2008). 182. Aoudjit, F. & Vuori, K. Matrix attachment regulates Fas-induced apoptosis in endothelial cells: a role for c-flip and implications for anoikis. J. Cell Biol. 152, 633–43 (2001). 183. Rosen, K., Shi, W., Calabretta, B. & Filmus, J. Cell detachment triggers p38 mitogen-activated protein kinase-dependent overexpression of Fas ligand. A novel mechanism of Anoikis of intestinal epithelial cells. J. Biol. Chem. 277, 46123–30 (2002). 184. Rosen, K. et al. Activated ras prevents downregulation of Bcl-X(L) triggered by detachment from the extracellular matrix: A mechanism of ras-induced

178

Bibliography resistance to anoikis in intestinal epithelial cells. J. Cell Biol. 149, 447–455 (2000). 185. Fukazawa, H., Noguchi, K., Masumi, a, Murakami, Y. & Uehara, Y. BimEL is an important determinant for induction of anoikis sensitivity by mitogen-activated protein/extracellular signal-regulated kinase kinase inhibitors. Mol Cancer Ther 3, 1281–1288 (2004). 186. Buchheit, C. L., Angarola, B. L., Steiner, A., Weigel, K. J. & Schafer, Z. T. Anoikis evasion in inflammatory breast cancer cells is mediated by Bim-EL sequestration. Cell Death Differ 22, 1275–1286 (2015). 187. Fisher, M. C. et al. Emerging fungal threats to animal, plant and ecosystem health. Nature 484, 186–94 (2012). 188. Brown, G. D., Denning, D. W. & Levitz, S. M. Tackling human fungal infections. Science (80-. ). 336, 647 (2012). 189. Brown, G. D. et al. Hidden killers: human fungal infections. Sci Transl Med 4, 165rv13 (2012). 190. Schneider, D. S. & Ayres, J. S. Two ways to survive infection: what resistance and tolerance can teach us about treating infectious diseases. Nat. Rev. Immunol. 8, 889–95 (2008). 191. Romani, L. Immunity to fungal infections. Nat Rev Immunol 11, 275–288 (2011). 192. Jouault, T. et al. Host responses to a versatile commensal: PAMPs and PRRs interplay leading to tolerance or infection by Candida albicans. Cell. Microbiol. 11, 1007–15 (2009). 193. Armstrong-James, D., Meintjes, G. & Brown, G. D. A neglected epidemic: fungal infections in HIV/AIDS. Trends Microbiol 22, 120–127 (2014). 194. Latge, J. P. The pathobiology of Aspergillus fumigatus. Trends Microbiol 9, 382– 389 (2001). 195. Ben-Ami,

R.,

Lewis,

R.

E.

&

Kontoyiannis,

D.

P. Enemy of

the

(immunosuppressed) state: an update on the pathogenesis of Aspergillus fumigatus infection. Br J Haematol 150, 406–417 (2010). 196. Bonnett, C. R., Cornish, E. J., Harmsen, A. G. & Burritt, J. B. Early neutrophil 179

Bibliography recruitment and aggregation in the murine lung inhibit germination of Aspergillus fumigatus conidia. Infect. Immun. 74, 6528–6539 (2006). 197. Paris, S. et al. Internalization of Aspergillus fumigatus conidia by epithelial and endothelial cells. Infect. Immun. 65, 1510–4 (1997). 198. Osherov, N. Interaction of the pathogenic mold Aspergillus fumigatus with lung epithelial cells. Front. Microbiol. 3, 346 (2012). 199. Wasylnka, J. A. & Moore, M. M. Aspergillus fumigatus conidia survive and germinate in acidic organelles of A549 epithelial cells. J. Cell Sci. 116, 1579–87 (2003). 200. Tomee, J. F. & Kauffman, H. F. Putative virulence factors of Aspergillus fumigatus. Clin Exp Allergy 30, 476–484 (2000). 201. Dagenais, T. R. & Keller, N. P. Pathogenesis of Aspergillus fumigatus in Invasive Aspergillosis. Clin Microbiol Rev 22, 447–465 (2009). 202. Knutsen, A. P. & Slavin, R. G. Allergic bronchopulmonary aspergillosis in asthma and cystic fibrosis. Clin. Dev. Immunol. 2011, 843763 (2011). 203. Verweij, P. E., Snelders, E., Kema, G. H. J., Mellado, E. & Melchers, W. J. G. Azole resistance in Aspergillus fumigatus: a side-effect of environmental fungicide use? Lancet. Infect. Dis. 9, 789–95 (2009). 204. Abad, A. et al. What makes Aspergillus fumigatus a successful pathogen? Genes and molecules involved in invasive aspergillosis. Rev Iberoam Micol 27, 155–182 (2010). 205. Choi, H. S., Shim, J. S., Kim, J.-A., Kang, S. W. & Kwon, H. J. Discovery of gliotoxin as a new small molecule targeting thioredoxin redox system. Biochem. Biophys. Res. Commun. 359, 523–8 (2007). 206. Dolan, S. K., O’Keeffe, G., Jones, G. W. & Doyle, S. Resistance is not futile: Gliotoxin biosynthesis, functionality and utility. Trends Microbiol. 23, 419–428 (2015). 207. Waksman, S. A. & Woodruff, H. B. Selective Antibiotic Action of Various Substances of Microbial Origin. J. Bacteriol. 44, 373–84 (1942). 208. Johnson, J. R., Bruce, W. F. & Dutcher, J. D. Gliotoxin, The Antibiotic Principle 180

Bibliography of Gliocladium fimbriatum. I. Production, Physical and Biological Properties 1. J. Am. Chem. Soc. 65, 2005–2009 (1943). 209. Bell, M. R., Johnson, J. R., Wildi, B. S. & Woodward, R. B. The Structure of Gliotoxin. J. Am. Chem. Soc. 80, 1001–1001 (1958). 210. Scharf, D. H. et al. Biosynthesis and function of gliotoxin in Aspergillus fumigatus. Appl Microbiol Biotechnol 93, 467–472 (2012). 211. Kwon-Chung, K. J. & Sugui, J. A. What do we know about the role of gliotoxin in the pathobiology of Aspergillus fumigatus? Med Mycol 47, S97–103 (2009). 212. Sugui, J. A. et al. Gliotoxin is a virulence factor of Aspergillus fumigatus: gliP deletion attenuates virulence in mice immunosuppressed with hydrocortisone. Eukaryot Cell 6, 1562–1569 (2007). 213. Owens, R. A., Hammel, S., Sheridan, K. J., Jones, G. W. & Doyle, S. A proteomic approach to investigating gene cluster expression and secondary metabolite functionality in Aspergillus fumigatus. PLoS One 9, e106942 (2014). 214. Gallagher, L. et al. The Aspergillus fumigatus protein GliK protects against oxidative stress and is essential for gliotoxin biosynthesis. Eukaryot. Cell 11, 1226–38 (2012). 215. Scharf, D. H. et al. Transannular disulfide formation in gliotoxin biosynthesis and its role in self-resistance of the human pathogen Aspergillus fumigatus. J. Am. Chem. Soc. 132, 10136–41 (2010). 216. Coleman, J. J., Ghosh, S., Okoli, I. & Mylonakis, E. Antifungal activity of microbial secondary metabolites. PLoS One 6, e25321 (2011). 217. Scharf, D. H., Brakhage, A. A. & Mukherjee, P. K. Gliotoxin- bane or boon? Environ. Microbiol. 18, 1096–1109 (2015). 218. Pahl, H., Krauß, B. & Vogt, M. The Immunosuppressive Fungal Metabolite Gliotoxin Specifically Inhibits Transcription Factor NF-KB. j. Exp. Med. 183, 12 (1996). 219. Coméra, C. et al. Gliotoxin from Aspergillus fumigatus affects phagocytosis and the organization of the actin cytoskeleton by distinct signalling pathways in human neutrophils. Microbes Infect. 9, 47–54 (2007).

181

Bibliography 220. Morton, C. O., Bouzani, M., Loeffler, J. & Rogers, T. R. Direct interaction studies between Aspergillus fumigatus and human immune cells; what have we learned about pathogenicity and host immunity? Front. Microbiol. 3, 413 (2012). 221. Stanzani, M. et al. Aspergillus fumigatus suppresses the human cellular immune response via gliotoxin-mediated apoptosis of monocytes. Blood 105, 2258–65 (2005). 222. Balloy, V. et al. Aspergillus fumigatus-induced interleukin-8 synthesis by respiratory epithelial cells is controlled by the phosphatidylinositol 3-kinase, p38 MAPK, and ERK1/2 pathways and not by the toll-like receptor-MyD88 pathway. J Biol Chem 283, 30513–30521 (2008). 223. Tsunawaki, S., Yoshida, L. S., Nishida, S., Kobayashi, T. & Shimoyama, T. Fungal metabolite gliotoxin inhibits assembly of the human respiratory burst NADPH oxidase. Infect. Immun. 72, 3373–82 (2004). 224. Spikes, S. et al. Gliotoxin production in Aspergillus fumigatus contributes to hostspecific differences in virulence. J Infect Dis 197, 479–486 (2008). 225. Chen, J. et al. Gliotoxin Inhibits Proliferation and Induces Apoptosis in Colorectal Cancer Cells. Mar. Drugs 13, 6259–73 (2015). 226. Pardo, J. et al. The mitochondrial protein Bak is pivotal for gliotoxin-induced apoptosis and a critical host factor of Aspergillus fumigatus virulence in mice. J Cell Biol 174, 509–519 (2006). 227. Skladny, H. et al. Specific Detection of Aspergillus Species in Blood and Bronchoalveolar Lavage Samples of Immunocompromised Patients by TwoStep PCR. J. Clin. Microbiol. 37, 3865–3871 (1999). 228. Schägger, H. Tricine-SDS-PAGE. Nat. Protoc. 1, 16–22 (2006). 229. Reid, T. et al. Rhotekin, a new putative target for Rho bearing homology to a serine/threonine kinase, PKN, and rhophilin in the rho-binding domain. - PubMed - NCBI. J Biol Chem 271, 13556–13560 (1996). 230. Koczorowska, M. M. et al. Fibroblast activation protein-α, a stromal cell surface protease, shapes key features of cancer associated fibroblasts through proteome and degradome alterations. Mol. Oncol. 10, 40–58 (2016).

182

Bibliography 231. Craig, R. & Beavis, R. C. TANDEM: matching proteins with tandem mass spectra. Bioinformatics 20, 1466–7 (2004). 232. Martens, L., Vandekerckhove, J. & Gevaert, K. DBToolkit: processing protein databases for peptide-centric proteomics. Bioinformatics 21, 3584–5 (2005). 233. Nilse, L. et al. Yeast membrane proteomics using leucine metabolic labelling: bioinformatic data processing and exemplary application to the ERintramembrane protease Ypf1. unpublished 234. Haug, G., Barth, H. & Aktories, K. Purification and Activity of the Rho ADP‐ Ribosylating Binary C2/C3 Toxin. 406, 117–127 (2006). 235. Ren, X. D. et al. Focal adhesion kinase suppresses Rho activity to promote focal adhesion turnover. J. Cell Sci. 113 ( Pt 2, 3673–8 (2000). 236. Sievers, F. et al. Fast, scalable generation of high-quality protein multiple sequence alignments using Clustal Omega. Mol. Syst. Biol. 7, 539 (2011). 237. Goujon, M. et al. A new bioinformatics analysis tools framework at EMBL-EBI. Nucleic Acids Res. 38, W695–W699 (2010). 238. Lei, K. & Davis, R. J. JNK phosphorylation of Bim-related members of the Bcl2 family induces Bax-dependent apoptosis. Proc. Natl. Acad. Sci. U. S. A. 100, 2432–2437 (2003). 239. Paquet, C., Schmitt, E., Beauchemin, M. & Bertrand, R. Activation of multidomain and BH3-only pro-apoptotic Bcl-2 family members in p53-defective cells. Apoptosis 9, 815–31 (2004). 240. Tournier, C. et al. Requirement of JNK for stress-induced activation of the cytochrome c-mediated death pathway. Science 288, 870–4 (2000). 241. Hongisto, V. et al. Lithium blocks the c-Jun stress response and protects neurons via its action on glycogen synthase kinase 3. Mol. Cell. Biol. 23, 6027–36 (2003). 242. Bain, J. et al. The selectivity of protein kinase inhibitors: a further update. Biochem. J. 408, 297–315 (2007). 243. Inuzuka, H. et al. SCF(FBW7) regulates cellular apoptosis by targeting MCL1 for ubiquitylation and destruction. Nature 471, 104–9 (2011).

183

Bibliography 244. Tournier, C. MKK7 is an essential component of the JNK signal transduction pathway activated by proinflammatory cytokines. Genes Dev. 15, 1419–1426 (2001). 245. Lin, A. et al. Identification of a dual specificity kinase that activates the Jun kinases and p38-Mpk2. Science 268, 286–90 (1995). 246. Haeusgen, W., Herdegen, T. & Waetzig, V. The bottleneck of JNK signaling: Molecular and functional characteristics of MKK4 and MKK7. Eur. J. Cell Biol. 90, 536–544 (2011). 247. Papa, S. et al. Insights into the structural basis of the GADD45beta-mediated inactivation of the JNK kinase, MKK7/JNKK2. J. Biol. Chem. 282, 19029–41 (2007). 248. Lee, D. E. et al. 7,3’,4'-Trihydroxyisoflavone, a metabolite of the soy isoflavone daidzein, suppresses ultraviolet B-induced skin cancer by targeting Cot and MKK4. J. Biol. Chem. 286, 14246–56 (2011). 249. Kim, N. et al. A Protoberberine derivative HWY336 selectively inhibits MKK4 and MKK7 in mammalian cells: the importance of activation loop on selectivity. PLoS One 9, e91037 (2014). 250. Fleming, Y. et al. Synergistic activation of stress-activated protein kinase 1/c-Jun N-terminal kinase (SAPK1/JNK) isoforms by mitogen-activated protein kinase kinase 4 (MKK4) and MKK7. Biochem. J. 352 Pt 1, 145–54 (2000). 251. Potin, S., Bertoglio, J. & Bréard, J. Involvement of a Rho-ROCK-JNK pathway in arsenic trioxide-induced apoptosis in chronic myelogenous leukemia cells. FEBS Lett. 581, 118–24 (2007). 252. Ohtsu, H. et al. Signal-crosstalk between Rho/ROCK and c-Jun NH2-terminal kinase mediates migration of vascular smooth muscle cells stimulated by angiotensin II. Arterioscler. Thromb. Vasc. Biol. 25, 1831–6 (2005). 253. Marinissen, M. J. et al. The Small GTP-Binding Protein RhoA Regulates c-Jun by a ROCK-JNK Signaling Axis. Mol. Cell 14, 29–41 (2004). 254. Dreikhausen, U. et al. Regulation by rho family GTPases of IL-1 receptor induced signaling: C3-like chimeric toxin and Clostridium difficile toxin B inhibit signaling

184

Bibliography pathways involved in IL-2 gene expression. Eur. J. Immunol. 31, 1610–9 (2001). 255. Orth, J., Schmidt, G. & Aktories, K. Bakterielle Toxine aktivieren GTPasen durch Deamidierung. Biospektrum 15, 501–503 (2009). 256. Spiering, D. & Hodgson, L. Dynamics of the Rho-family small GTPases in actin regulation and motility. Cell Adh. Migr. 5, 170–180 (2011). 257. Santos, M. F. et al. Rho proteins play a critical role in cell migration during the early phase of mucosal restitution. J. Clin. Invest. 100, 216–25 (1997). 258. McLean, G. W. et al. The role of focal-adhesion kinase in cancer — a new therapeutic opportunity. Nat. Rev. Cancer 5, 505–515 (2005). 259. Mor-Cohen, R., Rosenberg, N., Landau, M., Lahav, J. & Seligsohn, U. Specific cysteines in beta3 are involved in disulfide bond exchange-dependent and independent activation of alphaIIbbeta3. J. Biol. Chem. 283, 19235–44 (2008). 260. Tamura, T., Hato, T., Yamanouchi, J. & Fujita, S. Critical residues for ligand binding in blade 2 of the propeller domain of the integrin alphaIIb subunit. Thromb. Haemost. 91, 111–8 (2004). 261. Krokhin, O. V et al. Mass spectrometric based mapping of the disulfide bonding patterns of integrin alpha chains. Biochemistry 42, 12950–9 (2003). 262. Mould, A. P., Askari, J. A. & Humphries, M. J. Molecular basis of ligand recognition by integrin alpha 5beta 1. I. Specificity of ligand binding is determined by amino acid sequences in the second and third NH2-terminal repeats of the alpha subunit. J. Biol. Chem. 275, 20324–36 (2000). 263. Humphries, J. D., Byron, A. & Humphries, M. J. Integrin ligands at a glance. J. Cell Sci. 119, 3901–3903 (2006). 264. Byron, A. et al. Anti-integrin monoclonal antibodies. J. Cell Sci. 122, 4009–4011 (2009). 265. Reginato, M. J. et al. Integrins and EGFR coordinately regulate the pro-apoptotic protein Bim to prevent anoikis. Nat. Cell Biol. 5, 733–40 (2003). 266. Chiarugi, P. & Giannoni, E. Anoikis: A necessary death program for anchoragedependent cells. Biochem. Pharmacol. 76, 1352–1364 (2008).

185

Bibliography 267. Kong, W. et al. MicroRNA-155 is regulated by the transforming growth factor beta/Smad pathway and contributes to epithelial cell plasticity by targeting RhoA. Mol. Cell. Biol. 28, 6773–84 (2008). 268. O’Malley, B. W. Mechanisms of action of steroid hormones. N. Engl. J. Med. 284, 370–7 (1971). 269. Schmidt, S. et al. Glucocorticoid-induced apoptosis and glucocorticoid resistance: molecular mechanisms and clinical relevance. Cell Death Differ. 11, S45–55 (2004). 270. Ploner, C. et al. The BCL2 rheostat in glucocorticoid-induced apoptosis of acute lymphoblastic leukemia. Leukemia 22, 370–7 (2008). 271. Bouillet, P. et al. BH3-only Bcl-2 family member Bim is required for apoptosis of autoreactive thymocytes. Nature 415, 922–6 (2002). 272. Mombaerts, P. et al. RAG-1-deficient mice have no mature B and T lymphocytes. Cell 68, 869–77 (1992). 273. Spindler, V., Schlegel, N. & Waschke, J. Role of GTPases in control of microvascular permeability. Cardiovasc. Res. 87, 243–253 (2010). 274. Gorovoy, M. et al. RhoGDI-1 modulation of the activity of monomeric RhoGTPase RhoA regulates endothelial barrier function in mouse lungs. Circ. Res. 101, 50–8 (2007). 275. van Nieuw Amerongen, G. P. & van Hinsbergh, V. W. M. Endogenous RhoA Inhibitor Protects Endothelial Barrier. Circ. Res. 101, 7–9 (2007). 276. Bao, Z. et al. Evidence for the involvement of cofilin in Aspergillus fumigatus internalization into type II alveolar epithelial cells. BMC Microbiol. 15, 161 (2015). 277. Riento, K. & Ridley, A. J. Rocks: multifunctional kinases in cell behaviour. Nat. Rev. Mol. Cell Biol. 4, 446–56 (2003). 278. Feng, Y., LoGrasso, P. V, Defert, O. & Li, R. Rho Kinase (ROCK) Inhibitors and Their Therapeutic Potential. J. Med. Chem. 59, 2269–300 (2016).

186

Appendix

Appendix 15 Supplementary data

Appendix 1 | GT breaks disulfide bridges of Insulin. Recombinant human Insulin (5 µg) was incubated with the indicated concentrations of GT in PBS at 37°C overnight. GT was able to break the disulfide-bridged insulin chains at high concentration. DTT was used as a positive control to reduce S-S bonds. Higher molecular weight bands were visible in silver staining upon GT treatment.

187

Appendix

Appendix 2 | Suspension cells express integrins. Surface integrins were stained with labeled antibodies and quantified by FACS relative to an isotype control. (A) Integrin β1 was stained on the surface of murine adhesion (MEF, top) and suspension cells (BAF3, bottom). Both cell types expressed integrin β1. (B) Similarly, Integrin β3 was stained on human adhesion (BEAS-2B, top) and suspension cells (Jurkat, bottom). Representative histograms are shown.

188

Appendix

Appendix 3 | GT-induced anoikis differs from Poly-Hema induced cell death. GT-induced signaling was compared to the classic anoikis system, using Poly-Hema (PH) coated plates. MEK and ERK phosphorylation was initially high and lost, when adhesion after trypsinization was inhibited by PH. In contrast, this signal is initially low and increases in response to GT. Also the levels of the BH3-only protein Bad are regulated differently in both stimuli. Immunoblot generated by Andreas Geißler and retrieved from132 (Supplementary Figure S2).

189

Appendix

16 Erklärung 1. Ich erkläre hiermit, dass ich die vorliegende Arbeit ohne unzulässige Hilfe Dritter und ohne Benutzung anderer als der angegebenen Hilfsmittel angefertigt habe. Die aus anderen Quellen direkt oder indirekt übernommenen Daten und Konzepte sind unter Angabe der Quellen gekennzeichnet. Insbesondere habe ich hierfür nicht die entgeltliche Hilfe von Vermittlungs- beziehungsweise Beratungsdiensten (Promotionsberater oder anderer Personen) in Anspruch genommen. Niemand hat von mir unmittelbar oder mittelbar geldwerte Leistungen für Arbeiten erhalten, die im Zusammenhang mit dem Inhalt der vorgelegten Dissertation stehen.

2. Die Arbeit wurde bisher weder im In- noch im Ausland in gleicher oder ähnlicher Form einer anderen Prüfungsbehörde vorgelegt.

3. Die Bestimmungen der Promotionsordnung der Fakultät für Biologie sind mir bekannt, insbesondere weiß ich, dass ich vor Vollzug der Promotion zur Führung des Doktortitels nicht berechtigt bin.

Florian Haun

190

Appendix

17 Publications Related to this thesis: Geissler, A., Haun, F., Frank, D. O., Wieland, K., Simon, M. M., Idzko, M., … Borner, C. (2013). Apoptosis induced by the fungal pathogen gliotoxin requires a triple phosphorylation of Bim by JNK. Cell Death and Differentiation, 20(10), 1317– 1329. http://doi.org/10.1038/cdd.2013.78 Neumann, S., El Maadidi, S., Faletti, L., Haun, F., Labib, S., Schejtman, A., … Borner, C. (2015). How do viruses control mitochondria-mediated apoptosis? Virus Research, 209, 45–55. http://doi.org/10.1016/j.virusres.2015.02.026

Collaboration: De Ford, C., Heidersdorf, B., Haun, F., Murillo, R., Friedrich, T., Borner, C., & Merfort, I. (2016). The clerodane diterpene casearin J induces apoptosis of T-ALL cells through SERCA inhibition, oxidative stress, and interference with Notch1 signaling. Cell Death and Disease, 7(1), e2070. http://doi.org/10.1038/cddis.2015.413

Not related to this thesis: Okamoto, S.-I., Nakamura, T., Cieplak, P., Chan, S. F., Kalashnikova, E., Liao, L., … Lipton, S. A. (2014). S-nitrosylation-mediated redox transcriptional switch modulates neurogenesis and neuronal cell death. Cell Reports, 8(1), 217–28. http://doi.org/10.1016/j.celrep.2014.06.005 Haun, F., Nakamura, T., & Lipton, S. A. (2013). Dysfunctional mitochondrial dynamics in the pathophysiology of neurodegenerative diseases. Journal of Cell Death, 6(1), 27–35. http://doi.org/10.4137/JCD.S10847 Haun, F., Nakamura, T., Shiu, A. D., Cho, D.-H., Tsunemi, T., Holland, E. A., … Lipton, S. A. (2013). S-nitrosylation of dynamin-related protein 1 mediates mutant huntingtin-induced mitochondrial fragmentation and neuronal injury in Huntington’s disease. Antioxidants & Redox Signaling, 19(11), 1173–84. http://doi.org/10.1089/ars.2012.4928

Disclaimer: The data presented in chapter 9 is prepared for submission to Nature Cell Biology.

191

Appendix

18 Abbreviations A.f.

Aspergillus fumigatus

ABPA

Allergic Bronchopulmonary Aspergillo

AfD

A.f. diffusible product

Apaf-1

apoptotic protease activating factor 1

APS

Ammonium peroxide sulfate

ATF2

activating transcription factor 2

ATG

Autophagy related genes

ATP

Adenosine triphosphate

ASK1

Apoptosis signal-regulating kinase

Bak

Bcl-2 antagonist killer 1

BALf

broncho alveolar lavage fluid

Bax

Bcl-2 associated X protein

BC groove

c-terminus binding groove

Bcl-2

B-cell lymphoma-2

BEAS-2B

Bronchial epithelial cells

BH

Bcl-2 homology domains

Bid

Bcl-2 interacting domain death agonist

Bim

Bcl-2 interacting mediator of cell death

Bok

Bcl-2 related ovarian killer

BSA

Bovine serum albumin

C. elegans

Caenorhabditis elegans

CAD

caspase activated DNase

CARD

Caspase

activation

and

recruitment

domains cIAP

cellular inhibitors of apoptosis

CLRs

C-type lectin receptors

Complement IF

complement inhibitory factors

Conidial IF

Conidial inhibitory factor

CRP

c-reactive protein

CNFy

cytotoxic necrotizing factor y

DAMP

damage-associated molecular-pattern

192

Appendix DD

Death domain

DED

Death effector domain

DLC1

Dynein light chain 1

DMSO

Dimethyl sulfoxide

DR

Death receptor

DTT

Dithiothreitol

ECM

extracellular matrix

ECL

Enhanced chemiluminescence

EDTA

Ethylene diamine tetra acetic acid

EGF

epidermal growth factor

EGTA

Ethylene glycol tetra acetic acid

eIF

eukaryotic initiation factors

ELK

ETS domain-containing protein

ERAD

endoplasmic-reticulum-associated degradation

ERK

extracellular signal-regulated kinase-1/2

ERK

extracellular-signal-regulated kinases

FACS

Fluorescence activated cell sorting

FADD

Fas-associated protein with dead domain

FAK

Focal adhesion kinase

FAT

focal adhesion targeting domain

FLIP

FLICE-inhibitory protein

FOXO3

forkhead boxO 3

FRET

Fluorescence resonance energy transfer

GAP

GTPase activating protein

GDI

GDP dissociation inhibitors

GDP/GTP

Guanosindi/triphosphate

GEF

guanine nucleotide exchange factors

GPX4

glutathione peroxidase 4

Grb2

growth factor receptor-bound protein 2

GSH

glutathione

GSK3

glycogen synthase kinase 3

GT

Gliotoxin

HBG

hydrophobic binding groove

193

Appendix HC

hydrocortisone-acetate

HDAC

Histone deacetylase

HEPES

4-(2-hydroxyethyl)-1Piperazineethanesulfonic acid

IA

Invasive aspergillosis

IAP

inhibitor of apoptosis

IKK

IκB kinase

IL-3

Interleukin-3

IS

immunosuppressed

JNK

c-Jun N-terminal kinases 1/2

JNK

c-Jun n-terminal kinases

LUBAC

linear ubiquitin chain assembly complex

MAPK

mitogen activated protein kinases

MAPK

mitogen activated protein kinases

Mcl-1

myeloid cell leukemia 1

mDIA

mammalian homologue of Drosophila diaphanous

MEF

Mouse embryonic fibroblasts

MIDAS

metal ion-dependent adhesion site

MLC

myosin light chain

MLK

Mixed lineage kinase

MLKL

Mixed lineage kinase like

MOMP

Mitochondrial

outer

membrane

permeabilization MS

Mass spectrometry

MST1

mammalian sterile-20

MYPT1

myosin-binding

subunit

of

myosin

phosphatase 1 NCCD

Nomenclature Committee on Cell Death

Nec1

Necrostatin-1

NFκB

nuclear

factor

kappa-light-chain-

enhancer of activated B cells NT

Non-treated

OIP

Immunoprecipitation

194

Appendix P/S

Penicillin/ Streptomycin

PAGE

polyacrylamide gel electrophoresis

PAMPs

pathogen-associated molecular patterns

PARP

poly (ADP-ribose) polymerase

PB

Peripheral blood

PFA

Paraformaldehyde

PI

Propidiumiodide

PI

Propidium Iodide

PI3K

phosphoinositide 3-kinase

PIP2

phosphatidylinositol (3,4)-bisphosphate

PIP3

phosphatidylinositol (3,4,5)-triphosphate

PMSF

Phenylmethylsufonylfluoride

PRK

protein kinase C-related protein kinase

PRRs

pattern recognition receptors

PS

Phosphatidylserin

Puma

p53-upregulated modulator of apoptosis

Rag

recombination-activating gene

RHIM

RIP homotypic interaction motif

RIPK

receptor interacting protein kinase

ROCK

Rho-associated kinase

ROS

Reactive oxygen species

ROS

Reactive oxygen species

SDS

sodium dodecyl sulfate

SMAC

second mitochondria-derived activator of caspases

SOD

superoxide dismutase

SP

Surfactant protein

TAK1

TGF activated kinase 1

TLRs

Toll-like receptors

TNF

tumor necrosis factor

TNFR

tumor necrosis factor receptor

TRAIL

Tumor

Necrosis

Factor

Related

Apoptosis Inducing Ligand wt

Wild-type

195

Appendix

196